Opinion

The dispositive issue in this certified appeal is whether the Appellate Court properly concluded that the defendant, Larry Davis, had not been deprived of his right to a fair trial under the due process clause of the federal constitution1 by the joint trial of *20three legally unrelated informations.2 We conclude that, although the offenses charged in one of the three informations involved brutal and shocking conduct, the trial court’s thorough and proper jury instructions cured any risk of prejudice to the defendant. Accordingly, we affirm the judgment of the Appellate Court.

In connection with three separate incidents, the defendant was charged in three informations.3 The first information, docket number CR00-0490576, pertained to a shooting that occurred on September 28, 1998, in a parking lot located near Yale-New Haven Hospital, during which Victoria Standberry was wounded severely (Standberry information or Standberry case). In connection with that incident, the defendant was charged with assault in the first degree in violation of General Statutes §§ 53a-8 (a) and 53a-59 (a) (5), carrying a pistol without a permit in violation of General Statutes (Rev. to 1997) § 29-35, criminal possession of a firearm in violation of General Statutes (Rev. to 1997) § 53a-217, failure to appear in the first degree in violation of General Statutes § 53a-172 and, in a part B information, being a persistent dangerous felony offender in violation of General Statutes (Rev. to 1997) § 53a-40 (a) and (f). The second information, docket number CR030024537, pertained to an armed robbery of Lenwood E. Smith, Jr., that occurred on January 25, 2002 (Smith information or Smith case). In connection with that incident, the defendant was charged with robbery in *21the first degree in violation of General Statutes § 53a-134 (a) (4), larceny in the second degree in violation of General Statutes § 53a-123 (a) (3) and, in a part B information, being a persistent dangerous felony offender in violation of General Statutes § 53a-40 (a) and (h). The third information, docket number CR030024538, pertained to an armed robbery of Leonard Hughes that occurred on March 13, 2002 (Hughes information or Hughes case). In connection with that incident, the defendant was charged with burglary in the second degree in violation of General Statutes § 53a-102 (a) (2), robbery in the first degree in violation of § 53a-134 (a) (4) and larceny in the second degree in violation of § 53a-123 (a) (3).

Prior to trial, the state moved to consolidate, and the defendant moved to sever, the three separate informations. Additionally, the state moved to consolidate for trial a fourth information that charged the defendant with robbery in the first degree, criminal use of a firearm, risk of injury to a child and failure to appear in the first degree (fourth information). The defendant objected to the state’s motion for consolidation, claiming that the offenses charged in the fourth information and in the Standberry information were brutal and shocking in nature and, therefore, a joint trial would “impede the defendant’s constitutional right to a fair trial by an impartial jury.” In its memorandum of decision on the motion to consolidate, the trial court, Fasano, J., recognized that the offenses in the fourth information arose out of an incident in which the defendant allegedly had entered the home of an individual and had “demanded money at gunpoint and [had] threatened [that individual’s] children at gunpoint.” He therefore sustained the defendant’s objection with respect to the fourth information, concluding that an “armed threat to children could well fuel the prejudice of jurors against the defendant with respect to the other similar *22.crimes.” The trial court overruled the defendant’s objection with respect to the Standberry information, however, concluding that, although “the incident is obviously serious and involves violence, based upon the information before the court, it is not so brutal or shocking that its consolidation with the other matters would result in substantial injustice and prejudice beyond the curative power of the court’s instructions.” Accordingly, the trial court granted the state’s motion to consolidate for trial the Standberry, Smith and Hughes informations, but denied the state’s motion to consolidate the fourth information.

Throughout the course of the proceedings before the trial court, Licari, J., the defendant repeatedly renewed his objection to the order of consolidation, and moved to sever the Standberry, Smith and Hughes informations, claiming undue prejudice. The trial court, however, denied all of these motions, concluding that its detailed jury instructions were sufficient to cure any risk of prejudice to the defendant. During jury selection and throughout the trial, the trial court repeatedly and thoroughly instructed the jury that the Standberry, Smith and Hughes informations had been consolidated only for purposes of judicial efficiency, and that the evidence in each case must be considered separately and independently. Additionally, at trial, the state presented its evidence in each case chronologically and sequentially, beginning with the offenses charged in the Standberry information and ending with the offenses charged in the Hughes information.

The jury found the defendant guilty of all of the offenses charged in the Standberry and Smith informations, but not guilty of all of the offenses charged in the Hughes information. See also footnote 3 of this opinion. In a subsequent trial on the accompanying part B informations, the jury found the defendant guilty of two counts of being a persistent dangerous felony *23offender. The trial court rendered judgment in accordance with the jury’s verdict, and imposed a total effective sentence of eighty years imprisonment.

The defendant appealed from the trial court’s judgments of conviction to the Appellate Court claiming, inter alia, that the trial court improperly had granted the state’s motion to consolidate, and had denied the defendant’s motion to sever, the Standberry, Smith and Hughes informations for trial.4 The Appellate Court concluded that the trial court had not abused its discretion by consolidating the three informations for trial under the standards enunciated in State v. Boscarino, 204 Conn. 714, 720-21, 529 A.2d 1260 (1987), and its progeny. State v. Davis, 98 Conn. App. 608, 614-25, 911 A.2d 753 (2006). Specifically, the Appellate Court concluded that, “[e]ach case was sufficiently factually dissimilar so that the defendant was not exposed to potential prejudice from the jury”; id., 618; and that the offenses charged in the Standberry information were “not so brutal and shocking [when compared to the offenses charged in the Smith and Hughes informations so] as to result in unfair prejudice to the defendant.” Id., 622. Regardless, the Appellate Court concluded that the trial court’s “repeated and detailed jury instructions cured any prejudice” that may have flowed from the trial court’s order of consolidation. Id. Accordingly, the Appellate Court affirmed the judgments of the trial court; id., 638; and this certified appeal followed.

*24The jury reasonably could have found the following facts, as summarized in part by the Appellate Court, with respect to the offenses charged in the Standberry information. “In September, 1998, the first victim, [Standberry], had been introduced to the defendant by her best friend, Taraneisha Brown. Brown and the defendant were involved in a personal relationship. On September 27, 1998, Standberry asked Brown for payment toward a substantial debt owed by Brown. Brown replied that she would return Standberry’s telephone call but never did.

“The next day, the defendant received a telephone call in the afternoon and left work early. On the evening of September 28, 1998, Standberry parked her vehicle in the Pro Park parking lot located near Yale-New Haven Hospital (hospital), where she was employed in the food and nutrition department. Brown knew that Standberry parked in that particular lot when working at the hospital. Standberry left the hospital carrying a plate of food at approximately 9:25 p.m. and went to her vehicle. As she was placing the food in her vehicle, she observed an individual approach. She attempted to close her door, but it was forced open. The defendant came up to Standberry, said ‘revenge,’ and shot her several times before slowly walking away.” Id., 611. Despite severe physical injuries, Standberry was able to drive her vehicle, with the driver side door open and her injured leg hanging outside of the vehicle, to the entrance of the children’s hospital. An ambulance was summoned and Standberry was rushed to the emergency room, where she underwent several surgeries. Standberry testified that a cadaver bone was inserted in her shoulder to repair bone loss and nerve damage, and that two bullets remain in her body, one in her hip and one in her knee.5

*25The jury reasonably could have found the following facts, as summarized by the Appellate Court, with respect to the offenses charged in the Smith information. “The second victim, [Smith], was at a club in New Haven on January 25, 2002. After speaking with the defendant for approximately twenty minutes, he left at 2 a.m. The defendant stopped Smith in the parking lot and asked for a ride to Sheffield Street. Smith agreed, and the defendant and his friend entered Smith’s vehicle. After arriving, the defendant asked Smith to drive them to Carmel Street, where an individual known as ‘Mizzy’ owed him money. After Smith drove to the bottom of a hill, the defendant took out a gun and threatened him. Smith continued on to Carmel Street and parked. The defendant placed his gun against Smith’s head and demanded money. Smith gave the defendant his wallet and told him that he could get more from an automated teller machine. Smith drove to a nearby bank and, after parking, fled to a nearby gas station. Smith telephoned the police and showed them the bank parking lot where he had left his vehicle. The police recovered Smith’s vehicle approximately one week later.” Id., 612-13.

Lastly, “[a] summary of the evidence presented against the defendant with respect to [the Hughes information] is necessary for our discussion. There was evidence presented that Hughes was the superintendent of a building at 260 Dwight Street in New Haven. During the early morning of March 13, 2002, the defendant rang Hughes’ doorbell and said he was there to pick up items that an individual known as ‘Magnetic’ had left for him. These items included a motor vehicle, a safe, a bulletproof vest and 2.5 kilograms of cocaine. The defendant entered the apartment, pointed a gun at Hughes and ordered him to turn over the requested items. The defen*26dant took the keys to the motor vehicle and specifically asked for the cocaine. Hughes responded that there was no cocaine in the apartment. After being told to get on his knees, Hughes indicated that he would give the defendant the cocaine. The two men walked into a storage area, and Hughes managed to duck behind a steel door, escape through a window and flee to a nearby hotel. Hughes reported the incident to the police, who searched for the defendant, but were unable to locate him. Later that day, police officers recovered Hughes’ motor vehicle.” Id., 613-14.

On appeal to this court, the defendant renews the claim that he raised in the Appellate Court, namely, that the joint trial of the Standberry, Smith and Hughes informations resulted in substantial prejudice to the defendant because the evidence adduced in the Standberry case was brutal and shocking in nature and, therefore, inflamed the passions of the jurors beyond the curative powers of the trial court’s instructions.6 We disagree.

*27The principles that govern our review of a trial court’s ruling on a motion for joinder or a motion for severance are well established. Practice Book § 41-19 provides that, “[t]he judicial authority may, upon its own motion or the motion of any party, order that two or more *28informations, whether against the same defendant or different defendants, be tried together.” See also General Statutes § 54-57 (“[w]henever two or more cases are pending at the same time against the same party in the same court for offenses of the same character, counts for such offenses may be joined in one information unless the court orders otherwise”). “In deciding whether to sever informations joined for trial, the trial court enjoys broad discretion, which, in the absence of manifest abuse, an appellate court may not disturb. . . . The defendant bears a heavy burden of showing that the denial of severance resulted in substantial injustice, and that any resulting prejudice was beyond the curative power of the court’s instructions.” (Internal quotation marks omitted.) State v. Randolph, 284 Conn. 328, 337, 933 A.2d 1158 (2007).

“Substantial prejudice does not necessarily result from a denial of severance even [if the] evidence of one offense would not have been admissible at a separate trial involving the second offense. . . . Consolidation under such circumstances, however, may expose the defendant to potential prejudice for three reasons: First, when several charges have been made against the defendant, the jury may consider that a person charged with doing so many things is a bad [person] who must have done something, and may cumulate evidence against him .... Second, the jury may have used the evidence of one case to convict the defendant in another case even though that evidence would have been inadmissible at a separate trial. . . . [Third] joinder of cases that are factually similar but legally unconnected . . . presents] the . . . danger that a defendant will be subjected to the omnipresent risk . . . that although so much [of the evidence] as would be admissible upon any one of the charges might not [persuade the jury] of the accused’s guilt, the sum of it will convince them as to all. . . .

*29“Despite the existence of these risks, this court consistently has recognized a clear presumption in favor of joinder and against severance . . . and, therefore, absent an abuse of discretion . . . will not second guess the considered judgment of the trial court as to the joinder or severance of two or more charges. . . .

“The court’s discretion regarding joinder, however, is not unlimited; rather, that discretion must be exercised in a manner consistent with the defendant’s right to a fair trial. Consequently, we have identified several factors that a trial court should consider in deciding whether a severance may be necessary to avoid undue prejudice resulting from consolidation of multiple charges for trial. These factors include: (1) whether the charges involve discrete, easily distinguishable factual scenarios; (2) whether the crimes were of a violent nature or concerned brutal or shocking conduct on the defendant’s part; and (3) the duration and complexity of the trial. ... If any or all of these factors are present, a reviewing court must decide whether the trial court’s jury instructions cured any prejudice that might have occurred.” (Citation omitted; internal quotation marks omitted.) Id., 337-38; see also State v. Boscarino, supra, 204 Conn. 722-24.

We note that the defendant challenges the propriety of the Appellate Court’s conclusion that he was not prejudiced by the joint trial of the Standberry, Smith and Hughes informations on the basis of the second Boscarino factor only.7 Whether one or more offenses involve brutal or shocking conduct likely to arouse the passions of the jurors must be ascertained by comparing the relative levels of violence used to perpetrate the *30offenses charged in each information. See, e.g., State v. Ellis, 270 Conn. 337, 378, 852 A.2d 676 (2004) (defendant’s abuse of one victim “was substantially more egregious than his abuse of other two [victims]” and, therefore, joinder was improper under second Boscarino factor); State v. Horne, 215 Conn. 538, 549, 577 A.2d 694 (1990) (“[t]he brutality with which the assailant carried out the robbery and sexual assault in [one] case was very likely to have so aroused the passions of the jury that it interfered with their fair consideration of the other three cases” in which defendant was charged only with robbery); State v. Stevenson, 43 Conn. App. 680, 691, 686 A.2d 500 (1996) (“when all of the cases sought to be consolidated are brutal or shocking, they may be joined properly, if consolidation does not cause a high risk of one case being tainted by the unusually shocking ... or brutal nature of the other, or others”), cert. denied, 240 Conn. 920, 692 A.2d 817 (1997). Accordingly, to resolve the defendant’s claim, we must compare the evidence adduced in the Standberry case with the evidence adduced in the Smith and Hughes cases to ascertain whether the defendant’s conduct in the Standberry case was brutal and shocking in nature and, therefore, likely to have inflamed the passions of the jurors.

In the Standberry case, the victim, without warning, was shot multiple times at close range in an empty parking lot and suffered serious and extensive physical injuries, which required multiple surgeries to repair. The jury reasonably could have found that the shooting was not accidental or incidental to the completion of another crime, but, rather, was purposeful, premeditated and motivated by a desire for revenge. By contrast, in the Smith and Hughes cases, although physical violence was threatened, no injuries were inflicted and the victims escaped unharmed. Moreover, the threat of violence was incidental to the completion of another *31crime, namely, burglary, robbery and larceny, and was not, as in the Standberry case, the primary objective of the offense. Stated another way, in the Smith and Hughes cases, the threat of physical violence was a means to an end, whereas, in the Standberry case, ruthless physical violence was the end in and of itself. On the basis of the foregoing, we conclude that the defendant’s conduct in the Standberry case was significantly more brutal and shocking than his conduct in the Smith and Hughes cases, and that the level of violence used to perpetrate the crimes charged in the Standberry case was likely to have aroused the passions of the jurors. See State v. Ellis, supra, 270 Conn. 378; State v. Horne, supra, 215 Conn. 549; see also State v. Boscarino, supra, 204 Conn. 723 (“[w]e have acknowledged that evidence of a defendant’s brutal or shocking conduct in one case may compromise the jury’s ability to consider fairly the charges against him in other unrelated, but jointly tried cases”); State v. Silver, 139 Conn. 234, 240-41, 93 A.2d 154 (1952) (“Substantial injustice might result to a defendant where the evidence of one of the several crimes charged will show such brutality on his part that it is apt to arouse the passion of the jury against him to such an extent that they probably would not give fair consideration to the evidence relating to the other charges. Such a situation, however, is rare . . . .”).

The Appellate Court concluded, however, that the Standberry case did not involve brutal or shocking conduct. State v. Davis, supra, 98 Conn. App. 619-22. In arriving at this conclusion, the Appellate Court distinguished Ellis, Home and Boscarino by noting that those cases involved the crime of sexual assault, which “[s]hort of homicide ... is the ultimate violation of self ” (Emphasis in original; internal quotation marks omitted.) Id., 621, quoting State v. Ellis, supra, 270 Conn. 377; State v. Horne, supra, 215 Conn. 549-50; see also State v. Jennings, 216 Conn. 647, 659, 583 A.2d 915 *32(1990) (concluding that crimes of assault and kidnapping were not brutal and shocking in nature because “[t]he physical harm that was inflicted on the victim, although serious, was not disabling, and the element of sexual derangement present in Boscarino was absent”). We take this opportunity to clarify that the crimes of homicide and sexual assault are not the only crimes that are brutal and shocking under the second prong of the Boscarino test. Indeed, it is not the classification of the crime charged, but, rather, the level of violence and brutality used to perpetrate the crime that is dispositive of the court’s inquiry. See, e.g., State v. Ellis, supra, 377 (“Not all crimes of sexual assault . . . are equally brutal and shocking. . . . For example, although [s]exual assaults in the first degree can be characterized as brutal . . . [s]ome . . . evince a greater degree of brutality or shocking behavior than others. The question then becomes whether one of the sexual assault crimes ... is so brutal and shocking when compared with the other, that a jury, even with proper instructions, could not treat them separately.” [Citation omitted; internal quotation marks omitted.]); State v. Herring, 210 Conn. 78, 97, 554 A.2d 686 (“[w]hile any murder involves violent and upsetting circumstances, it would be unrealistic to assume that any and all such deaths would inevitably be so ‘brutal and shocking’ that a jury, with proper instructions to treat each killing separately, would necessarily be prejudiced by a joint trial”), cert. denied, 492 U.S. 912, 109 S. Ct. 3230, 106 L. Ed. 2d 579 (1989). It is beyond dispute that an unprovoked, ruthless and violent assault that results in life threatening physical injuries, as in the Standberry case, may be so brutal and shocking, especially when compared to the less violent offenses charged in a legally unrelated information, that the ability of the jury to deliberate objectively and dispassionately on the jointly tried cases may be impaired.

*33Having concluded that the defendant’s conduct in the Standbeny case was brutal and shocking when compared to his conduct in the Smith and Hughes cases, we next consider whether the trial court’s jury instructions were sufficient to cure the risk of prejudice to the defendant. See State v. Atkinson, 235 Conn. 748, 766-67, 670 A.2d 276 (1996) (“in cases in which the likelihood of prejudice is not overwhelming . . . [the trial court’s] curative instructions may tip the balance in favor of a finding that the defendant’s right to a fair trial has been preserved” [internal quotation marks omitted]). The record reflects that, during voir dire, the trial court instructed each potential panel of jurors that the Standberry, Smith and Hughes informations had been consolidated for trial only for purposes of judicial efficiency, and that the evidence presented in each case must be considered separately and independently.8 The *34trial court repeated these instructions after the jury was impaneled,9 in its final charge10 and after the jury had *35returned its verdicts on the offenses charged in the Smith and Standberry informations.* 11 We conclude that the trial court’s thorough, explicit and proper jury instructions cured the risk of prejudice to the defendant and, therefore, preserved the jury’s ability to consider fairly and impartially the offenses charged in the jointly *36tried cases.12 See State v. Rivera, 260 Conn. 486, 493, 798 A.2d 958 (2002) (“ [although we might disagree with the trial court’s conclusion that the two cases were not brutal or shocking, we cannot say, as a reviewing court, that the trial court’s conclusion, coupled with proper and adequate jury instructions, constituted an abuse of discretion”).

Our conclusion on this point is buttressed by the sequence in which the jury rendered its verdicts, as well as the substance of those verdicts. After the Standberry, Smith and Hughes cases had been submitted to the jury for deliberation, the jury indicated that it had reached a unanimous verdict with respect to each of the offenses charged in the Smith case. The jury found the defendant guilty of those offenses, the trial court accepted the jury’s verdicts and the jury resumed its deliberations on the offenses charged in the Standberry and Hughes cases. Thereafter, the jury indicated that it had reached a unanimous verdict with respect to each of the offenses charged in the Standberry case. The jury found the defendant guilty of those offenses, the trial court accepted the jury’s verdicts and the jury resumed its *37deliberations on the offenses charged in the Hughes case. Finally, the jury indicated that it had reached a unanimous verdict with respect to each of the offenses charged in the Hughes case. The jury found the defendant not guilty of those offenses and the trial court accepted the jury’s verdicts. We conclude that the sequential order in which the jury rendered its verdicts reveals that the jury considered the evidence adduced in each separately and independently in accordance with the trial court’s instructions. Moreover, by acquitting the defendant of all of the offenses charged in the Hughes case, the jury evidently was able to keep the three cases separate and did not blindly condemn the defendant on the basis of the evidence adduced in the Standberry case. See State v. Atkinson, supra, 235 Conn. 766 (“by returning a verdict of not guilty on the charge of possession of a weapon in a correctional institution . . . the jury evidently was able to separate the two cases and did not blindly condemn the defendant on his participation in the murder”); State v. Rodriguez, 91 Conn. App. 112, 120-21, 881 A.2d 371 (“Although the jury found the defendant guilty of all the counts of burglary, attempt to commit burglary, larceny and criminal trespass that it considered, it found the defendant not guilty of one count of breach of the peace in the second degree. That acquittal demonstrated that the jury was able to consider each count separately and, therefore, was not confused or prejudiced against the defendant.”), cert. denied, 276 Conn. 909, 886 A.2d 423 (2005).13 Accordingly, we conclude that the defendant *38was not prejudiced by the joint trial of the Standberry, Smith and Hughes informations.

The judgment of the Appellate Court is affirmed.

In this opinion VERTEFEUILLE and ZARELLA, Js., concurred.

joins, concurring. I agree with the majority’s conclusion that the judgment of conviction of the defendant, Larry Davis, must be affirmed, despite the improper consolidation of the charges in one case pertaining to the victim Victoria Standberry (Standberry case) with the charges in two other cases pertaining to the victims Lenwood E. Smith, Jr., and Leonard Hughes (Smith and Hughes cases). I write separately, however, for two reasons.

First, I take this, opportunity to revisit the liberal presumption in favor of joinder that has been applied under our case law. In my view, the uniform application of such a presumption, irrespective of whether the evidence in each case would be cross admissible, cannot be reconciled with the well established rule barring evidence of other crimes as inherently prejudicial unless that evidence would be legally relevant to the case on some other basis. I therefore propose a refinement to the rule, under which joinder is presumptively favored only when the substantive evidence would be cross admissible in independent prosecutions; in the absence of such cross admissibility, prejudice is presumed and joinder will be proper only when the Boscarino factors1 demonstrate that the risk of prejudice is substantially reduced. This approach is both consistent with our treatment of uncharged misconduct evidence and recognizes the reality that a defendant faces when *39forced to defend against multiple charges—the presumption that he has a bad character and a propensity for criminal behavior, which in turn is likely to influence improperly the jury’s deliberations.

Second, I take this opportunity to clarify the standard that the reviewing court must apply in considering a challenge to a trial court’s decision granting joinder or denying severance. Our case law has tended to conflate what should be a two part inquiry. Consistent with the reviewing court’s role in examining any other claim of nonconstitutional error, it is clear that there are two questions that must be addressed in the affirmative before a defendant is entitled to a new trial: First, did the trial court abuse its discretion in granting joinder or denying severance? Second, did that decision result in harmful error?

In accordance with those inquiries, in the present case, I would conclude that the trial court abused its discretion in consolidating the cases because it was evident at the outset that the more brutal conduct alleged in the Standberry case might compromise the jury’s ability to consider the charges in the other two cases. I nonetheless would conclude that the defendant has failed to sustain his burden of proving that this impropriety constituted harmful error because the totality of the record provides us with a fair assurance that the improper consolidation did not affect the outcome.

I

Our case law on joinder consistently has recognized a long-standing rule of evidence, under which admission of evidence of other crimes categorically is proscribed unless that evidence is legally relevant to some other issue in the case. See, e.g., State v. Pollitt, 205 Conn. 61, 69, 530 A.2d 155 (1987); State v. Boscarino, 204 Conn. 714, 721-22, 529 A.2d 1260 (1987); State v. Jonas, 169 Conn. 566, 572-73, 363 A.2d 1378 (1975), cert. *40denied, 424 U.S. 923, 96 S. Ct. 1132, 47 L. Ed. 2d 331 (1976); State v. Oliver, 161 Conn. 348, 361, 288 A.2d 81 (1971); State v. Silver, 139 Conn. 234, 240-41, 93 A.2d 154 (1952). Although one could argue that the exceptions tend to swallow the rule; see Conn. Code Evid. § 4-5 (b);2 this court has not wavered from firm application of this rule when an exception is not demonstrated. See, e.g., State v. Randolph, 284 Conn. 328, 356-62, 933 A.2d 1158 (2007). Despite our rigid adherence to this rule of evidence, the court often has stated that a universal presumption in favor of joinder and against severance applies, irrespective of whether the evidence pertaining to one offense would be admissible as proof of the other offenses joined. See, e.g., id., 338; State v. McKenzie-Adams, 281 Conn. 486, 521, 915 A.2d 822, cert. denied, 552 U.S. 888, 128 S. Ct. 248, 169 L. Ed. 2d 148 (2007); State v. Ellis, 270 Conn. 337, 375, 852 A.2d 676 (2004); State v. Delgado, 243 Conn. 523, 532, 707 A.2d 1 (1998); State v. Chance, 236 Conn. 31, 38, 671 A.2d 323 (1996), ’writ of habeas corpus denied sub nom. Chance v. Kupec, United States District Court, Docket No. 3:96CV2204, 1998 U.S. Dist. LEXIS 18997 (D. Conn. November 17, 1998); State v. Jones, 234 Conn. 324, 344, 662 A.2d 1199 (1995). Such a uniform presumption seems to me to be in direct conflict with our adherence to fundamental rules of evidence.

*41As the District of Columbia Circuit Court of Appeals explained in Drew v. United States, 331 F.2d 85, 89-90 (D.C. Cir. 1964), a case often cited in this court’s case law on joinder:3 “It is a principle of long standing in our law that evidence of one crime is inadmissible to prove disposition to commit crime, from which the jury may infer that the defendant committed the crime charged. Since the likelihood that juries will make such an improper inference is high, courts presume prejudice and exclude evidence of other crimes unless that evidence can be admitted for some substantial, legitimate purpose. The same dangers appear to exist when two crimes arejoinedfor trial, and the same principles of prophylaxis are applicable,4

“Evidence of other crimes is admissible when relevant to (1) motive, (2) intent, (3) the absence of mistake or accident, (4) a common scheme or plan embracing the commission of two or more crimes so related to each other that proof of the one tends to establish the other, and (5) the identity of the person charged with the commission of the crime on trial. When the evidence is relevant and important to one of these five issues, it is generally conceded that the prejudicial effect may be outweighed by the probative value.

*42“If, then, under the rules relating to other crimes, the evidence of each of the crimes on trial would be admissible in a separate trial for the other, the possibility of ‘criminal propensity’ prejudice would be in no way enlarged by the fact of joinder. When, for example, the two crimes arose out of a continuing transaction or the same set of events, the evidence would be independently admissible in separate trials. Similarly, if the facts surrounding the two or more crimes on trial show that there is a reasonable probability that the same person committed both crimes due to the concurrence of unusual and distinctive facts relating to the manner in which the crimes were committed, the evidence of one would be admissible in the trial of the other to prove identity. In such cases the prejudice that might result from the jury’s hearing the evidence of the other crime in a joint trial would be no different from that possible in separate trials.” (Emphasis added.)

That court and others, therefore, have endorsed a rule that “requires a severance of offenses that are purportedly of the ‘same or similar character’ unless evidence of the joined offenses would be mutually admissible in separate trials or, if not, unless the evidence is sufficiently ‘simple and distinct’ to mitigate the dangers otherwise created by such a joinder.” (Emphasis added.) United States v. Halper, 590 F.2d 422, 431 (2d Cir. 1978). Indeed, commentators generally have been critical of joinder in the absence of cross admissibility. See United States v. Foutz, 540 F.2d 733, 738 n.4 (4th Cir. 1976) (citing federal treatises and American Bar Association’s Project on Minimum Standards for Criminal Justice suggesting that joinder of similar offenses, when evidence is not cross admissible, generally should not be permitted). The proposition in United States v. Halper, supra, 431, that the prejudice must be “mitigated” to make joinder proper implicitly acknowl*43edges a presumption in such cases that must be overcome.

The tension between this reasoning and this court’s uniform presumption in favor of joinder has grown appreciably since this court’s decision in State v. King, 187 Conn. 292, 445 A.2d 901 (1982), wherein the court reconciled a conflict between the statute authorizing joinder of offenses of the “same character”; General Statutes § 54-57; and the rule of practice permitting joinder of dissimilar offenses; Practice Book § 41-19 (formerly § 829); in favor of the rule of practice. State v. King, supra, 296-98 (concluding that rule of practice permissibly expanded circumstances in which joinder could be proper because it regulates court procedure and does not infringe on any substantive right); see J. Bruckmann, G. Nash & J. Katz, Connecticut Criminal Caselaw Handbook (1989) p. 142 (“[t]he conflict arising from the failure [of § 54-57] to provide for consolidation of dissimilar cases was settled in [King] ... in favor of [the rule of practice]”). Although King recognized that the more liberal rule of practice on joinder “should be read in light of [now Practice Book § 41-18] permitting severance if prejudice may result”; (internal quotation marks omitted) State v. King, supra, 296; it opened the door to joinder of many more offenses that might not be cross admissible and hence increased the risk of prejudice to the defendant.

In my view, it also is significant that the presumption in favor of joinder is based on the rationale that it fosters judicial economy. See State v. Ellis, supra, 270 Conn. 375. This court has failed to acknowledge in King or subsequent cases, however, that the interests favoring joinder may weigh differently when both the offenses are dissimilar and the evidence is not cross admissible. As one treatise observed: “The argument for joinder is most persuasive when the offenses are based upon the same act or criminal transaction, since *44it seems unduly inefficient to require the state to resolve the same issues at numerous trials. Commentators have been generally critical, however, of the joinder of offenses which are unrelated, since the need to prove each offense with separate evidence and witnesses eliminates any real savings in time or efficiency which might otherwise be provided by a single trial.” A. Spinella, Connecticut Criminal Procedure (1985) p. 416. Moreover, the only realistic benefit to the defendant from joinder—receiving a speedier trial because of the reduced pressure on criminal dockets—is likely to be outweighed in dissimilar cases by the possibility that joinder may infringe on the defendant’s right to pursue different defenses or his right to testify as to one offense and not the other, and even in similar cases where the evidence is not cross admissible by the possibility that joinder may instill the presumption that the defendant has a bad character and a propensity for criminal behavior.

As the Fourth Circuit Court of Appeals noted: “[Although it is true that the . . . [r]ules of [c]riminal [procedure [were] designed to promote economy and efficiency and to avoid a multiplicity of trials . . . we are of the strong opinion that the consideration of one’s constitutional right to a fair trial cannot be reduced to a cost/benefit analysis. Thus, while we are concerned with judicial economy and efficiency, our overriding concern in an instance such as this is that [the] jury consider only relevant and competent evidence bearing on the issue of guilt or innocence for each individually charged crime separately and distinctly from the other.” (Citation omitted; internal quotation marks omitted.) United States v. Isom, 138 Fed. Appx. 574, 581 (4th Cir. 2005), cert. denied, 546 U.S. 1124, 126 S. Ct. 1103, 163 L. Ed. 2d 915 (2006); see also id. (concluding that trial court nonetheless did not abuse its discretion in denying defendant’s motion to sever because “[a]ny prejudice *45[he] suffered by having the two robbery charges joined into one trial is substantially mitigated by the fact that much of the evidence of one robbery would be admissible in the other”). Accordingly, I would instruct the trial courts that the presumption in favor of joinder is limited to cases wherein there is cross admissibility of substantive evidence. When the evidence would not be cross admissible, trial courts should presume prejudice and grant joinder only when the risk of prejudice appears to be “substantially reduced.”5 Drew v. United States, supra, 331 F.2d 91.

II

I next turn to the question of the proper standard to be applied by the reviewing court in deciding whether joinder was proper. In my view, the court improperly has conflated what should be a two-pronged inquiry: (1) whether the trial court abused its discretion; and (2) whether that impropriety constituted harmful error. We apply this rubric to every other claim of nonconstitu*46tional error, and I see no reason to do otherwise in our review of a claim of improper joinder.

In this court’s early case law on joinder, the court recognized that the reviewing court’s determination as to whether the trial court abused its discretion necessarily must be based on the evidence before the court when ruling on the motion: “Where from the nature of the case it appears that a joint trial will probably be prejudicial to the rights of one or more of the parties, a separate trial should be granted when properly requested. The discretion of the court is necessarily exercised before the trial begins, and with reference to the situation as it then appears .... The controlling question is whether it appears that a joint trial will probably result in substantial injustice. It is not necessarily a ground for granting a separate trial that evidence will be admissible against one of the accused which is not admissible against another. . . . When the existence of such evidence is relied on as a ground for a motion for separate trials, the character of the evidence and its effect upon the defense intended to be made should be stated, so that the court may be in a position to determine the probability of substantial injustice being done to the moving party from a joint trial. It does not appear from the record that the trial court was so advised in this case, and on that ground alone it is impossible to say that the court abused its discretion in denying the [defendant’s] motion.”6 (Emphasis added; internal quotation marks omitted.) *47State v. Castelli, 92 Conn. 58, 63, 101 A. 476 (1917); accord State v. Holup, 167 Conn. 240, 245, 355 A.2d 119 (1974) (“Because a preliminary motion for separate trials obviously must be decided before the actual trial, the merits of the motion can be determined only on the basis of whether at that time it appears that injustice is likely to result unless separate trials are held. It is for this reason that in support of such a motion the court must be fully informed of any and all circumstances which indicate that justice to the parties requires separate trials.”).

Indeed, were the reviewing court not to limit its initial abuse of discretion determination to the evidence then before the trial court, there would be a “grave danger of mistrials from causes which were unknown to the trial court at the time when it was required to decide the question.” State v. Castelli, supra, 92 Conn. 65. The trial court’s rulings on such motions usually are predicated on the face of the charging document and whatever information is provided to the court regarding evidence to be adduced at trial. Therefore, the reviewing court necessarily must base its determination as to whether the trial court abused its discretion by looking to the state of the record at the time the trial court acted, not to the fully developed record after trial.7 See, *48e.g., State v. Oliver, supra, 161 Conn. 360-62 (finding “no error in the preliminary ruling by the court which denied the defendant’s motion for separate trials on the two counts [of indecent assault] in the light of the circumstances as they were then before the court” when it was only after evidence had been adduced at trial that it became apparent that one victim’s identification had been tainted; ultimately reversing judgment and remanding for new trial on independent ground that there was substantial likelihood that inadmissible identification had been substantial factor in jury’s verdict of guilt as to other offense); State v. Klein, 97 Conn. 321, 324-25, 116 A. 596 (1922) (“In the present case two confessions or statements were offered, besides other evidence of lesser consequence, and admitted against one or two of the accused. None of this evidence, except as to one confession, was known to the court at the time these motions [for separate trials] were heard and decided. . . . With the fact of the single confession before it, we cannot say that the court abused its discretion in denying these motions.” [Citation omitted.]).

Although the dispositive question is prejudice, that question is viewed from a predictive perspective when considering whether the trial court had abused its discretion when acting on the motion to join or sever, but is viewed from a fully informed perspective when determining whether improper joinder was harmful: “The test for the trial court is whether substantial injustice is likely to result unless a separate trial be *49accorded. The test for this court is whether the denial of the motion for a separate trial has resulted in substantial injustice to the accused.” (Emphasis added.) State v. Klein, supra, 97 Conn. 324.

In State v. Boscarino, supra, 204 Conn. 714, the court applied three factors when determining whether joinder had been proper: (1) whether the “factual similarities . . . [although] insufficient to make the evidence in each case substantively admissible at the trial of the others, were significant enough to impair the defendant’s right to the jury’s fair and independent consideration of the evidence in each case”; id., 723; (2) whether “[t]he prejudicial impact of joinder in these cases was exacerbated by the violent nature of the crimes with which the defendant was charged . . . [giving] the state the opportunity to present the jury with the intimate details of each of these offenses, an opportunity that would have been unavailable if the cases had been tried separately”; id.; and (3) whether “[t]he duration and complexity of the trial also enhanced the likelihood that the jury would weigh the evidence against the defendant cumulatively, rather than independently in each case.” Id. Boscarino did not make clear, however, that these factors are pertinent in determining both whether the trial court’s decision was proper in the first instance and whether the defendant received a fair trial despite improper joinder.8 See State v. Herring, 210 *50Conn. 78, 96 n.16, 554 A.2d 686 (1989) (“Several of the factors that we stressed in [.Boscarino] require hindsight in determining whether the defendant received a fair trial. While it may be relevant to consider whether the defendant raised the question of prejudice at trial or requested appropriate curative instructions, the effect of a denial of severance may be difficult to predict in advance of the actual testimony at trial.”), cert. denied, 492 U.S. 912, 109 S. Ct. 3230, 106 L. Ed. 2d 579 (1989). The court’s failure to make this point exacerbated a tendency in our case law to conflate the requisite inquiry, considering the totality of the record at the conclusion of trial, rather than first considering the evidence as it appeared to the trial court at the time it ruled on the motion. See, e.g., State v. Horne, 215 Conn. 538, 548-51, 577 A.2d 694 (1990) (determining that evidence adduced at trial exposed defendant to prejudice and then considering whether trial court’s instruction was adequate to mitigate that prejudice); State v. Bell, 188 Conn. 406, 411, 450 A.2d 356 (1982) (“[s]ince the state was able to present the evidence in an orderly manner and since it appears that the jury was not confused and was able to consider the evidence as to each charge separately and distinctly, it is clear that the trial court did not abuse its discretion in permitting a single trial”); State v. Jonas, supra, 169 Conn. 571 (“our analysis must focus on the nature of the evidence produced at the trial”).

Whether the trial court abused its discretion in joining offenses in the first instance, however, cannot resolve the question of whether such an improper ruling requires reversal. It is a well settled rule that, “[w]hen a trial error in a criminal case does not involve a constitutional violation the burden is on the defendant to *51demonstrate the harmfulness of the court’s error.” (Internal quotation marks omitted.) State v. Iban C., 275 Conn. 624, 640, 881 A.2d 1005 (2005); accord State v. Smith, 280 Conn. 285, 307, 907 A.2d 73 (2006) (“When an improper evidentiary ruling is not constitutional in nature, the defendant bears the burden of demonstrating that the error was harmful. . . . We have recently stated that a nonconstitutional error is harmless when an appellate court has a fair assurance that the error did not substantially affect the verdict.” [Citation omitted; internal quotation marks omitted.]). I can find no rationale in our case law to justify a different analytical framework for improper joinder than that which we apply in every other type of nonconstitutional impropriety, wherein the court engages in a two part inquiry to determine first whether there was an impropriety, and second whether the impropriety was harmful error in light of the record as a whole.9 Indeed, the only time the court does not apply a two step analysis is in the limited class of claims that constitutes structural error, wherein the claim “is not susceptible to a harmless error analysis . . . .” (Internal quotation marks omitted.) State v. Brown, 279 Conn. 493, 505, 903 A.2d 169 (2006).

*52Instructive in this regard is this court’s decision in State v. Hamilton, 228 Conn. 234, 235, 636 A.2d 760 (1994), wherein the issue was “the proper standard for appellate review of a trial court’s denial of a motion for continuance to retain private counsel.” The court noted: “The determination of whether to grant a request for a continuance is within the discretion of the trial court, and will not be disturbed on appeal absent an abuse of discretion.” (Internal quotation marks omitted.) Id., 239. The court then explained the proper standard of review as follows: “In appellate review of matters of continuances, federal and state courts have identified multiple factors that appropriately may enter into the trial court’s exercise of its discretion. Although the applicable factors cannot be exhaustively catalogued, they generally fall into two categories. One set of factors focuses on the facts of record before the trial court at the time when it rendered its decision. From this perspective, courts have considered matters such as: the timeliness of the request for continuance; the likely length of the delay; the age and complexity of the case; the granting of other continuances in the past; the impact of delay on the litigants, witnesses, opposing counsel and the court; the perceived legitimacy of the reasons proffered in support of the request; the defendant’s personal responsibility for the timing of the request; the likelihood that the denial would substantially impair the defendant’s ability to defend himself; the availability of other, adequately equipped and prepared counsel to try the case; and the adequacy of the representation already being afforded to the defendant. . . . Another set of factors has included, as part of the inquiry into a possible abuse of discretion, a consideration of the prejudice that the defendant actually suffered by reason of the denial of the motion for continuance. . . . For purposes of assessing actual prejudice, the focus is on the adequacy of the defen*53dant’s legal representation subsequent to the trial court’s ruling, as distinguished from its likely adequacy as determined by the trial court at the time of its ruling on the motion for continuance. . . .

“Although our past rulings on this question have not been entirely consistent . . . the decision of the Appellate Court in this case affords us the opportunity to clarify that an appellate court should limit its assessment of the reasonableness of the trial court’s exercise of its discretion to a consideration of those factors, on the record, that were presented to the trial court, or of which that court was aware, at the time of its ruling on the motion for a continuance.

“In the event that the trial court acted unreasonably in denying a continuance, the reviewing court must also engage in harmless error analysis. If a claim on appeal is nonconstitutional in nature, ‘the burden of establishing that harm resulted from a trial court error rests on the appellant. . . . When a continuance has been requested to obtain new counsel after a trial has begun, the defendant must show, on appeal, that the improper denial of the motion demonstrably prejudiced his ability to defend himself.’ ” (Citations omitted; emphasis added.) Id., 240-42. In a footnote appended to the last sentence quoted from Hamilton, the court noted: “We recognize that this analysis has, in the past, been effectuated under the rubric of ‘abuse of discretion’ . . . and that this has led to inconsistency in the application of the standard. For purposes of clarification, therefore, we now term the assessment of actual prejudice to the defendant’s case as ‘harmless error analysis.’ ” (Citations omitted.) Id., 242 n.4.

I would make clear, as the court did in Hamilton, that the reviewing court must not conflate the distinct inquiries relating to abuse of discretion and harmful error. The reviewing court first must determine whether *54the trial court abused its discretion in light of the information before the court when it ruled on the motion. If there was such an abuse of discretion, the reviewing court then must determine whether the defendant has established that, in light of the totality of evidence at trial and the trial court’s subsequent instructions to the jury, the impropriety constituted harmful error.

Ill

Applying this analytical framework to the present case, I would conclude that the trial court abused its discretion in denying the defendant’s motion to sever because the evidence in the cases was not cross admissible and application of the Boscarino factors would not have demonstrated that the risk of prejudice from joinder of the more brutal allegations in the Standberry case with those in the other cases was “substantially reduced.” Upon review of the totality of the record, however, I further would conclude that the defendant has failed to prove that this impropriety was harmful error.

At the outset, I note that the pertinent inquiry is whether the evidence regarding the improperly joined case, involving the defendant’s premeditated “revenge” shooting of Standberry, compromised the jury’s ability to consider fairly the charges against him of first degree robbery and larceny in the Smith case. I agree with the defendant that the trial court’s remedial instructions were not sufficient to mitigate the prejudice caused by the improper joinder in this case. The general instructions for the jury to keep the evidence in each case separate, which must be given whenever legally unrelated cases are joined in a single trial, did not address the particular prejudice at issue here. See State v. Horne, supra, 215 Conn. 552-53 (“The trial court also asked the jury somehow to ignore the obviously inflammatory nature and impact of the sexual assault. It failed, *55however, to instruct the jury about the difficulties of this assignment at the outset of the trial, despite its initial agreement to the defendant’s request that the court make such an instruction. The government should not have the windfall of having the jury be influenced by evidence against a defendant which, as a matter of law, they should not consider but which they cannot put out of their minds.” [Internal quotation marks omitted.]); see also State v. Boscarino, supra, 204 Conn. 724-25 (“[A] curative instruction is not inevitably sufficient to overcome the prejudicial impact of [inadmissible other crimes] evidence. ... In the circumstances of these cases, we conclude that even the trial court’s apt and thorough admonitions could not mitigate the potential for prejudice wrought by the joinder of the cases against the defendant.” [Citation omitted; internal quotation marks omitted.]). Harmful error analysis, however, does not begin and end with this limited aspect of the record. Cf., e.g., State v. Thompson, 266 Conn. 440, 456, 832 A.2d 626 (2003) (concluding that, although admission of certain testimony was abuse of discretion, it was harmless error because, inter alia, state’s attorney did not emphasize or rely upon testimony during closing argument and there was significant other evidence of defendant’s guilt); State v. Hafford, 252 Conn. 274, 297, 746 A.2d 150 (“[t]his court has held in a number of cases that when there is independent overwhelming evidence of guilt, a constitutional error would be rendered harmless beyond a reasonable doubt” [internal quotation marks omitted]), cert. denied, 531 U.S. 855, 121 S. Ct. 136, 148 L. Ed. 2d 89 (2000). The dispositive question in harmful error analysis is whether we have a fair assurance that the defendant received a fair trial. See State v. Smith, supra, 280 Conn. 307.

The defendant in the present case did not contest that Smith had been robbed, nor did he suggest that *56Smith had a motive to lie. Rather, the sole defense was misidentification. The evidence regarding Smith’s identification, however, was quite strong. Smith had ample opportunity to observe the defendant that night. Smith testified that, when he first had been approached by the defendant at a club on the night of the robbery, the defendant looked familiar from previous contact some time earlier. Smith and the defendant spoke for approximately twenty minutes before leaving the club. After they left the club together in Smith’s car, the defendant sat in the passenger’s seat directly next to Smith for approximately twenty more minutes. Smith testified that the defendant had referred to himself as “Lord Devine” or “Devine,” and the defendant stipulated to the fact that he had been known by those names since the late 1990s. Approximately two months after the robbery, Smith identified the defendant from a photographic array. Although the defendant attempted to attack inconsistencies between Smith’s statement to the police on the night of the robbery and his subsequent statement two months after the robbery, those inconsistencies either were unrelated to identification or were insignificant matters, such as whether Smith had said that the defendant had “a braid” or “braids” of hair sticking up. Thus, the state’s evidence in the Smith case was extremely strong and was not impeached in any substantive manner.

Finally, I note, as did the majority, that the jury deliberated and rendered its verdict of guilty in the Smith case prior to its deliberations in the Standberry case. This chronology further diminished the likelihood that the more brutal facts in the Standberry case played a role in the jury’s deliberations in the Smith case. Given the totality of the record, I would conclude that the court has a fair assurance that the improper joinder did not affect the verdict in the Smith case. Accordingly, I respectfully concur in the judgment.