OPINION OF THE COURT

In a tragic incident occurring on Thanksgiving Day, November 25, 1976, appellant Robert Torsney, a New York City police officer, shot and killed a 15-year-old Black youth. He was indicted and charged with second degree murder. At his trial, held in November, 1977, one year after the incident, Torsney interposed as a defense a lack of criminal responsibility for his conduct on the date of the crime charged stemming from a mental disease or defect diagnosed as psychomotor epilepsy. On November 30, 1977, the jury returned a verdict of not guilty by reason of mental disease or defect. (Penal Law, § 30.05.) '

. On that same date, Torsney was ordered committed to the custody of the Commissioner of the Department of Mental Hygiene, petitioner-appellant, pursuant to CPL 330.20 (subd 1) . After initial commitment to Mid-Hudson Psychiatric Center, Torsney was transferred to Creedmoor Psychiatric Center on March 3, 1978. Shortly thereafter, a recommendation was made to the Director of Creedmoor that inasmuch as Torsney was not dangerous or mentally ill, he should be released. On May 19, 1978, a special release committee was convened for the purpose of examining Torsney to determine his suitability for release. After the special release committee agreed with the staffs findings and recommendations, in which the then Director of Creedmoor, Dr. Werner, concurred, the findings and recommendations were reported to the commissioner. The commissioner thereupon convened an independent review panel, which also concurred in the recommendation.

Thereafter, on July 20, 1978, pursuant to CPL 330.20 (subd 2) , the commissioner petitioned the committing court for an order discharging Torsney from his custody. The court ordered a full evidentiary hearing and at the conclusion of which sustained the petition and ordered Torsney released pursuant *671to CPL 330.20 (subd 3)1 upon the following conditions: (1) that he not be permitted to carry a gun; (2) that he not be a police officer or peace officer; (3) that he continue as an outpatient at Creedmoor for a period of five years on conditions to be imposed as determined by Creedmoor Psychiatric Center; and (4) if within five years after the conditional release the court should determine after a hearing that for the safety of Mr. Torsney or of others the conditional release should be revoked, the court must recommit him.2 On cross appeals taken by the District Attorney and by the commissioner and Torsney, the Appellate Division reversed and ordered that Torsney be recommitted to the custody of the Commissioner of Mental Hygiene.

On this appeal, two issues are presented for our review. The threshold issue is whether the Appellate Division properly construed the standard for release of persons held in the custody of the Commissioner of the Department of Mental Hygiene pursuant to CPL 330.20. Contingent upon resolution of this issue is the second issue: namely, whether evaluated under the proper standard for release the weight of the credible evidence presented at the hearing requires the detainee’s continued confinement, discharge or release on condition.

We begin our analysis of the threshold issue with CPL 330.20 (subd 1), which provides in pertinent part: "Upon rendition of a verdict of accquittal by reason of mental disease or defect, the court must order the defendant to be committed to the custody of the commissioner of mental, hygiene to be placed in an appropriate institution in the state department of *672mental hygiene.” In recently sustaining the constitutionality of the automatic commitment of persons acquitted by reason of mental disease or defect, this court observed that "[a]n individual who has committed an act of violence, and has thus demonstrated his dangerousness, and who has successfully asserted an insanity defense, may quite properly be treated somewhat differently from other individuals who, although they may in fact be potentially equally dangerous as a result of mental problems, have not yet so vehemently demonstrated their dangerousness by violent antisocial behavior.” (People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d 334, 338; see, also, People v Lally, 19 NY2d 27; People ex rel. Peabody v Chanler, 133 App Div 159, affd 196 NY 525 [sustaining constitutionality of predecessor statutes to CPL 330.20].) For this reason, persons acquitted under CPL 330.20 may be viewed as an "exceptional class” justifying commitment to the custody of the Commissioner of Mental Hygiene without a prior hearing to determine their mental condition on the date of acquittal.

The purpose of automatic commitment after acquittal of a crime by reason of mental disease or defect is a narrow one: to determine the mental condition of the person committed on the date of acquittal. Notwithstanding the literal terms of CPL 330.20, automatic commitment of persons acquitted of crimes by reason of mental disease or defect is constitutionally permissible only for a reasonable period of time — that is, sufficient time to permit an examination and report as to the detainee’s sanity. (People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d, at p 339, supra; People v Lally, 19 NY2d, at p 33, supra; People v McNelly, 83 Misc 2d 262, 266; Lee v Kolb, 449 F Supp 1368, 1379.) Moreover, a prompt hearing must be held on this issue, a hearing at which a person committed under CPL 330.20 — like any other involuntarily committed patient seeking release from custody — is entitled to a jury trial. (Mental Hygiene Law, § 9.35; People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d, at p 339, supra; People v Lally, 19 NY2d, at pp 34-35, supra.) To permit commitment of a person without such a hearing to determine his present mental condition and dangerousness would constitute a deprivation of due process and equal protection of the law. (See Bolton v Harris, 395 F2d 642, 650-652; Wilson v State, 259 Ind 375; People v McQuillan, 392 Mich *673511; State v Krol, 68 NJ 236; State ex rel. Kovach v Schubert, 64 Wis 2d 612, app dsmd 419 US 1117, cert den 419 US 1130.)

The purpose of this hearing, characterized as a civil proceeding, is to determine whether the detainee "may be discharged or released on condition without danger to himself or others.” (CPL 330.20, subd 3.) The Appellate Division equated the standard of dangerousness under CPL 330.20 with the standard of dangerousness for involuntary civil commitment under the Mental Hygiene Law. Specifically, the court incorporated by reference the standard for involuntary admission contained in section 9.37 of the Mental Hygiene Law, which provides for the admission of

"any person who, in the opinion of the director of community services or his designee, has a mental illness for which immediate inpatient care and treatment in a hospital is appropriate and which is likely to result in serious harm to himself or others; 'likelihood of serious harm’ shall mean:

"1. substantial risk of physical harm to himself as manifested by threats of or attempts at suicide or serious bodily harm or other conduct demonstrating that he is dangerous to himself, or

"2. a substantial risk of physical harm to other persons as manifested by homicidal or other violent behavior by which others are placed in reasonable fear or serious physical harm.”

In adopting this definitional standard, however, the Appellate Division, although noting that this section requires as a condition precedent to application of the "likelihood of serious harm” test that the patient suffer a mental illness requiring immediate in-patient care and treatment, expressly held that a detainee under CPL 330.20 need not satisfy this criteria to preclude his discharge. In that court’s view, "[a] dangerous detainee [can] be confined even if not mentally ill or in need of immediate treatment, under CPL 330.20.” (66 AD2d 281, 288, n 6.) We disagree.

Unquestionably, the unique status of persons acquitted of a crime by reason of mental disease or defect permits the State to treat this "exceptional class” somewhat differently from other individuals believed suffering from mental disease or defect. Because such persons have demonstrated past antisocial behavior, the State is permitted to engage in what may be broadly termed a presumption that the causative mental *674illness or defect continues beyond the date of the criminal conduct to the date of acquittal. (People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d, at p 340, supra.) Of course, the flaw in viewing this principle as a true presumption is readily apparent: namely, the possibility that years may have intervened between commission of the crime charged and the date of acquittal. The significance of this contingency is further buttressed when consideration is given to the fact that in New York a lack of criminal responsibility by reason of mental illness or defect is a defense rather than an affirmative defense. (Penal Law, § 30.05.) Thus, an acquittal predicated on this ground indicates only that the People failed to prove beyond a reasonable doubt that the defendant was criminally responsible for his act and not that the jury found by a preponderance of the evidence that he was suffering from a mental disease or defect on the date of the commission of the crime charged. (Wilson v State, 259 Ind, at p 385, supra; see, also, Note, Commitment of Persons Acquitted by Reason of Insanity: The Example of the District of Columbia, 74 Col L Rev 733, 744-750.)

Thus, it may be said more accurately that the presumption flowing from an acquittal by reason of mental disease or defect does not presume that a person so acquitted presently suffers from mental disease or defect, but, rather, posits merely that having raised this defense and having previously engaged in antisocial behavior, he has demonstrated sufficient grounds for further examination to determine his present need for treatment and confinement as opposed to immediate release. (See People v McQuillan, 392 Mich 511, 527, supra.) It is for this purpose only — a prompt examination and report as to sanity — that such a person may be automatically committed to the custody of the Commissioner of Mental Hygiene upon acquittal. (People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d, at p 339, supra; People v Lally, 19 NY2d, at p 33, supra; Wilson v State, 259 Ind, at p 386, supra; State v Krol, 68 NJ, at p 256, supra; People v McQuillan, 392 Mich, at p 528, supra; see, generally, Hamann, The Confinement and Release of Persons Acquitted by Reason of Insanity, 4 Harv J Leg 55, 65-67.)

Beyond automatic commitment of persons found not guilty by reason of mental disease or defect for a reasonable period to determine their present sanity, justification for distinctions in treatment between persons involuntarily committed under *675the Mental Hygiene Law and persons committed under CPL 330.20 draws impermissibly thin. Clearly, the Mental Hygiene Law requires for involuntary civil commitment that the patient be mentally ill and in need of immediate in-patient treatment. Absent such a finding, a dangerous propensity is, in itself, insufficient to permit involuntary commitment. In our view, to permit commitment of a person acquitted of a crime by reason of mental disease or defect on a lesser substantive predicate would run afoul of the constitutional guarantees to due process and equal protection of the laws.

In Jackson v Indiana (406 US 715), a mentally defective deaf mute charged with two separate counts of robbery was found incompetent to stand trial and committed to the Indiana Department of Mental Health until certified sane. Counsel for Jackson contended that in seeking to commit Jackson to a mental institution indefinitely, the State should have been required to invoke the standards and procedures for commitment of "feeble-minded” or mentally ill persons provided under Indiana law. In accepting this contention, the court held "that by subjecting Jackson to a more lenient commitment standard and to a more stringent standard of release than those generally applicable to all others not charged with offenses, and by thus condemning him in effect to permanent institutionalization without the showing required for commitment or the opportunity for release * * * Indiana deprived petitioner [Jackson] of equal protection of the laws under the Fourteenth Amendment.” (406 US, at p 730, supra [footnote omitted].)3

In reaching this conclusion, the court relied on its prior decision in Baxstrom v Herold (383 US 107), in which the court held that a State prisoner may not be civilly committed at the expiration of his prison sentence on the finding of a Surrogate when the State affords all other civilly committed persons a jury trial on the issue of their sanity. (383 US, at p 110; see United States ex rel. Schuster v Herold, 410 F2d 1071^^^L.den 396 US 847.) "If criminal conviction and *676imposition of sentence are insufficient to justify less procedural and substantive protection against indefinite commitment than that generally available to all others,” reasoned the court, "the mere filing of criminal charges surely cannot suffice.” (Jackson v Indiana, 406 US, at p 724, supra.) The court then acknowledged that the Baxstrom principle had been extended to commitment upon acquittal by reason of mental disease or defect. (406 US, at p 724, supra, citing Bolton v Harris, 395 F2d 642, supra; Cameron v Mullen, 387 F2d 193; People v Lally, 19 NY2d 27, supra.)

In the present case, we deal with an individual falling conceptually between the petitioners in Baxstrom and Jackson: Torsney has been charged with a crime, has stood trial, and has been acquitted by reason of mental disease or defect. That equal protection mandates that he be afforded the same procedural rights governing his release from custody as any other involuntarily committed person is settled. Similarly, in our view, appellants’ petition for release must be measured by the same substantive standards governing involuntary civil commitment of any other individual. (See Wilson v State, 259 Ind, at p 386, supra.) Thus, we interpret CPL 330.20 as requiring a detainee’s release unless it is found that he is presently dangerous to himself or others by reason of a mental disease or defect. (See Overholser v O’Beirne, 302 F2d 852, 854-855; Starr v United States, 264 F2d 377, 382-383; Newton v Brooks, 246 Ore 484, 490; Mills v State, 256 A2d 752, 757, n 4 [Del]; State ex rel. Leeb v Wilson, 27 Ohio App 2d 1, 2-3.)

As in the case of involuntary civil commitment, it is the need for immediate in-patient treatment which justifies deprivation of liberty upon a theory of dangerousness. (See Gold-stein and Katz, Dangerousness and Mental Illness Some Observations on the Decision to Release Persons Acquitted by Reason of Insanity, 70 Yale LJ 225, 237-239.) A showing that a detainee may be viewed as dangerous because of the prior criminal act for which he was acquitted is insufficient without more to preclude his discharge or release on condition. Distinctions in treatment of persons alleged to be mentally ill cannot rest solely upon the existence of a prior criminal charge or even conviction. (Association of the Bar, City of New York, Special Committee on the Study of the Commitment Procedures and the Law Relating to Incompetents, Second Report, Mental Illness, Due Process and the Criminal Defen*677dant 1.) As Judge Fahy observed in Ragsdale v Overholser (281 F2d 943, 949, 950 [concurring opn]) some 19 years ago: "It is by no means clear that society can continue to deprive a person of liberty by attributing to a jury’s doubt about his mental condition, which led to his acquittal and mandatory commitment, any and all evil or criminal propensities he may be thought to have, and to keep him in confinement because of them. This would transform the hospital into a penitentiary where one could be held indefinitely for no convicted offense, and this even though the offense of which he was previously acquitted because of doubt as to his sanity might not have been one of the more serious felonies.” In the instant case, although faced with the most serious of felonies we must nonetheless be guided by this principle.

With this constitutionally mandated interpretation of CPL 330.20 in mind, we proceed to the second issue in this case: whether the weight of the credible evidence presented at the hearing requires continued confinement or release of the detainee on condition. Torsney committed the crime for which he was acquitted by reason of mental illness or defect on November 25, 1976.4 Two psychiatrists examined him for the purpose of determining whether he lacked criminal responsibility for his act. Dr. Daniel Schwartz, Torsney’s witness, concluded that as a result of mental defect, which he diagnosed as "psychosis associated with epilepsy”, Torsney did not at the time of the offense "know or appreciate the nature, consequence and wrongfulness of his act.” Dr. Herbert Spiegel, who examined Torsney on behalf of the People, concluded quite to the contrary that although Torsney was "prone to hysterical dissociation under stress”, he was not psychotic and was criminally responsible for his act. Paralleling these findings were those of two psychologists who examined Torsney: Dr. Joy Roy voiced the opinion that Torsney’s conduct on the date of the crime was consonant with the anxiety and bewilderment which one would expect to see in an epileptic equivalent. Dr. Florence Schumer opined, however, that although Torsney "might well have responded in an impulsive and volatile manner to anticipated threat”, little in the record indicated "cognitive pathology, psychoses, or bizarre or disorganized intellective or emotional processes.”

Of course, the question whether on this proof the People *678failed to demonstrate beyond a reasonable doubt that Torsney was criminally responsible for his act is not now material. Suffice it to say that upon the testimony produced at trial Torsney was acquitted on November 30, 1977 of the crime charged by reason of mental illness or defect. Thereupon he was admitted to Mid-Hudson Psychiatric Center on December 6, 1977, pursuant to the automatic commitment procedure contained in CPL 330.20. Torsney was examined by Dr. Mark Vandenbergh, who concluded that his reasoning and judgment was unimpaired and his insight normal. Dr. Vandenbergh found no evidence of mental disorder. Similarly, Dr. Alan Halpern, a psychologist at Mid-Hudson, who examined Torsney on December 7, determined that he suffered neither from a psychosis, a psychopathic disorder, nor from organic damage. Dr. Halpern did recommend, however, that because Torsney was a "somewhat rigid, impulsive individual”, he was not fit for duty as a beat police officer. Neither doctor believed Torsney was in need of any in-patient treatment.

Shortly thereafter, on January 6, 1978, Torsney was again examined by Dr. Vandenbergh, who found that appellant had "made an excellent adjustment to his present hospitalization.” More specifically, Dr. Vandenbergh observed that "neither before or after the offense [had Torsney] shown any signs of epilepsy”. Dr. Vandenbergh also noted that notwithstanding the neurotic emotional conflicts described in the psychological evaluations produced at trial, Torsney was symptom-free. In Dr. Vandenbergh’s opinion, he showed no symptoms of anxiety, ate and slept well and related normally to people around him, confirming his diagnostic impression that Torsney was without mental disorder.

The final examination of Torsney at Mid-Hudson Psychiatric Center occurred on February 9, 1978. This examination was conducted by Drs. Vandenbergh and Easwara Bhoothalingom, who agreed that Torsney showed no signs of mental disorder. Dr. Bhoothalingom, the Forensic Unit Chief at Mid-Hudson, noted on February 28, just prior to Torsney’s transfer to Creedmoor Psychiatric Center, that Torsney had "been cooperative with no evidence of any symptoms or signs during his stay here which point towards any potential for dangerous behavior.”

On March 3, 1978, Torsney was transferred from Mid-Hudson to Creedmoor, where he was examined upon admission by Dr. John McKnight. Dr. McKnight’s report indicated that *679Torsney did not appear either dangerous or suicidal and that his physical condition was within "normal limits”. In April, 1978, Dr. Deborah Kaiser, Chief of Service of the Forensic Unit at Creedmoor, recommended to the Director of Creed-moor, Dr. William Werner, that Torsney be released. Accordingly, on May 19, a special release committee was convened by Dr. Werner, consisting of Drs. Joseph Sklar, Sablón Dartigue and John McKnight. It was the unanimous opinion of this committee that Torsney was not suffering from a psychosis, psychopathic disorder or organic damage and that he was not dangerous to himself or others and should be released. The committee did, however, recommend that he not be given duty as a beat police officer.

Upon receipt of this report, Dr. Werner examined Torsney and notified Dr. John Wright, Assistant Commissioner of the Department of Mental Hygiene, that in his opinion Torsney had "recovered from his psychotic episode [showed], no evidence of mental illness and [was] not considered dangerous to himself or others”. Dr. Werner concurred in the recommendation that he be discharged. At the direction of the commissioner, an independent review panel was convened to examine Torsney. This panel, consisting of two independent psychiatrists, Drs. A. Louis McGarry and Stephen Rachlin, together with a certified social worker, Rosalind Silver, found, in a report dated July 6, 1978, that Torsney "can safely be released from the hospital, without danger to himself or others. His history reveals no dangerous behavior other than the instant offense, nor has he subsequently shown any inability to control impulses, any inappropriate display of hostility, nor any behavior which might be considered threatening or dangerous. He voices no intentions to do harm to anyone and is not now mentally ill, having fully recovered from the entity which led to his [acquittal] by reason of mental disease or defect.”

Upon the submission by the Commissioner of Mental Hygiene of an application pursuant to CPL 330.20 requesting Torsney’s discharge or release on condition, the court, at the request of the District Attorney, appointed two independent forensic psychiatrists to examine Torsney and submit recommendations to the court. The first of these psychiatrists, Dr. Milton Hollar, concluded, based upon his examination of Torsney and a review of his records, that he was not psychotic; that no evidence existed to indicate that he was then dangerous to himself or others; and that no justification *680existed for continued incarceration on psychiatric grounds. The second court-appointed psychiatrist, Dr. Stanley Portnow, similarly concluded in his report: "Mr. Torsney is neither a lunatic or a psychopath and is not a menace to others or to himself. In my opinion, there is nothing about his present mental condition which warrants his continued incarceration, and he appears to be capable of returning to a productive and useful place in society.” In view of the prior history of the case and what Dr. Portnow termed Torsney’s "questionable impulse control and inability to deal with stress”, Dr. Portnow recommended that he be gradually discharged over several weeks to ensure his satisfactory return to the community.

After receiving these reports, the court directed that a hearing be held pursuant to CPL 330.20 (subd 3). At the hearing, which ran a full nine days and at which some 18 WitheSses testified, agreement was unanimous: Torsney was not presently suffering from a mental illness or defect and Was hot a danger to himself or others. Every opinion elicited recommended that he be discharged or released on condition.

The first Witness called by the Attorney-General on behalf of the Department of Mental Hygiene, Dr. Kaiser, testified that during Torsney’s stay at Creedmoor he "exhibited no signs nor symptoms of mental illness. His reality testing [was] good, his judgment [was] good. He [showed] no impulsive behavior. He most certainly [showed] none of the usual signs that one might think of associated with mental illness such as hallucinations, delusions, poor judgment, cognitant problems, any such things. He [showed] no evidence of any such seizure disorder. In short, this person [presented] himself to us and [did] so since his admission as being normal.” Dr. Kaiser further testified that she had observed no evidence of aggressive behavior or assaultiveness on Torsney’s part. Nor had she witnessed any racial prejudice. Moreover, when questioned as to whether she had considered the effect which stress in the outside world might work on Torsney, particularly pressures flowing from the tragic prior incident, Dr. Kaiser testified that those problems had been discussed in depth and that she did not anticipate that Torsney "would become dangerous considering the stresses he may face.”

Dr. Sklar, a member of the special release committee, was called by the District Attorney and testified that upon his examination of Torsney he found him "responsive, goal-directed, mildly depressed” and that he showed "no evidence at *681that time of an emotional disorder that required in-patient treatment from a clinical point of view.” The witness also testified that in reaching his conclusion that he was not dangerous and should be released, he took into consideration the everyday stresses of life in the community outside the hospital. Dr. Dartigue the second member of the special release committee,5 also called by the District Attorney, reiterated his agreement with this recommendation and acknowledged that he similarly had considered stress outside the hospital. More specifically, when asked what stress outside the hospital he had considered, Dr. Dartigue responded that he had considered whether Torsney, if put in the same condition as November 25, 1976, would behave the same way. Having considered this possibility, Dr. Dartigue nonetheless recommended release.

Also testifying at the hearing were the two court-appointed independent psychiatrists. Dr. Hollar confirmed the view recommended in his report: namely, that Torsney was not psychotic and not dangerous to himself or others. Although frankly admitting a lack of infallible prescience when queried whether he could say that Mr. Torsney "in no way” had the capacity to repeat the tragic incident of November, 1976, Dr. Hollar explained that in his judgment based upon the control he had demonstrated in the hospital, psychological examinations and his own examination, Torsney was sufficiently in control of himself and he felt free in recommending his discharge.

The second independent psychiatrist, Dr. Portnow, similarly confirmed his findings and recommendation made in the previously submitted report to the court. His testimony supported that of all the other expert witnesses called: Torsney was in his judgment neither mentally ill nor dangerous to himself or others. In view of the prior incident in his history, Dr. Port-now again voiced his opinion that Torsney be released through some form of controlled release program. He acknowledged, however, that the program recommended by Dr. Kaiser, which would have required Torsney to report once a week to an outpatient clinic at the hospital so that he could be continuously evaluated for five years, was in his opinion a suitable program.

Finally, when queried as to his statement in his report *682concerning Torsney’s "questionable impulse control”, Dr. Port-now explained that although he felt that Torsney was "an impulsive somewhat explosive individual”, he was not mentally ill. Dr. Portnow further clarified that "[ejxplosive personality is the designation given to a personality as opposed to a mental disease or defect which is described primarily as, as the instantaneous, almost without warning, explosion of effect of temperment as [sic] a particular situation.” When asked to provide a reason for his finding that Torsney had an impulsive or explosive personality, Dr. Portnow stated: "Mr. Torsney told me himself that he considered himself to be somewhat impulsive, he was an impulsive buyer, he was a man of action, given to action without thinking first and generally those are the characteristics which one would apply to an impulsive or explosive personality.”

Integrating the testimony elicited at the hearing with the proper standard for discharge or release on condition under CPL 330.20, we conclude that the weight of the credible evidence mandates that the order of the hearing court be reinstated. One thing is abundantly clear: every opinion offered at the hearing substantiated Torsney’s claim that he is neither suffering from a mental illness or defect nor dangerous to himself or others. We reject as constitutionally suspect the interpretation of CPL 330.20 given by the Appellate Division, which would permit a detainee’s continued institutionalization on a vague concept of dangerousness unrelated to mental illness or defect and for which no immediate in-patient treatment is required. As Chief Judge Bazelon cautiously observed: " '[DJangerousness’ is a many splendored thing. Unless muzzled by discriminating analysis, it is likely to weigh against nominally competing considerations the way a wolf weighs against a sheep in the same scales: even if the sheep is heavier when weighed separately, somehow the wolf always prevails when the two are weighed together. Keeping dangerousness on a taut leash is especially difficult where there is danger of murder, since the danger is admittedly grave and since its improbability, which theoretically discounts its gravity, is exceedingly difficult to quantify.” (Covington v Harris, 419 F2d 617, 627.)

An individual’s liberty cannot be deprived by "warehousing” him in a mental institution when he is not suffering from a mental illness or defect and in no need of in-patient care *683and treatment6 on a ground which amounts to a presumption of a dangerous propensity flowing from, as in this case, an isolated, albeit tragic, incident occurring years ago. Were this the standard for involuntary institutionalization, logical extension would require that anyone convicted of a violent crime should upon completion of his sentence similarly be required to demonstrate that he or she is not dangerous before his release into the community. (See Matter of Williams, 157 F Supp 871, 876, affd sub nom. Overholser v Williams, 252 F2d 629.) Whatever its label, confinement on a showing of mere propensity amounts to nothing more than preventive detention, a concept foreign to our constitutional order. (Cross v Harris, 418 F2d 1095, 1102; see, generally, Dershowitz, Preventive Confinement: A Suggested Framework for Constitutional Analysis, 51 Tex L Rev 1277.)

Of course, this is not to say that artificial, and perhaps theoretically implausible, lines must be drawn between concepts of mental disease or defect and other impairments of judgment. As the Appellate Division noted: "[Psychiatrists continually redefine their understanding of what a mental disease or defect is.” (66 AD2d 281, 295.) This difficulty has gained added significance since the adoption of the Durham rule as a means of determining a defendant’s criminal responsibility for his conduct pursuant to which mental illness has been construed to encompass psychopathic and sociopathic personalities. (See Goldstein and Katz, Dangerousness and Mental Illness Some Observations on the Decision to Release Persons Acquitted by Reason of Insanity, 70 Yale LJ 225, 238.) The friction between the operation of a perhaps theoretically subjective rule permitting acquittals on something less than uniform notions of mental disease or defect and the cry of due process for objective criteria upon which to predicate continued institutionalization is amply illustrated in the case at bar. Perhaps resolution of this conflict lies in re-evaluation of the scope and applicability of the insanity defense.7 (See, gener*684ally, Burt, Of Mad Dogs and Scientists: The Perils of the "Criminal — Insane”, 123 U Pa L Rev 258, 280-285; German and Singer, Punishing the Not Guilty: Hospitalization of Persons Acquitted by Reason of Insanity, 29 Rutgers L Rev 1011, 1056-1058.) Given the unsettled state of the law in this area, we do not suggest that in a proper case a ’’personality disorder” which resulted in a finding of dangerousness and which required treatment would not be sufficient to preclude a detainee’s discharge or release on condition. Suffice it to say, on the present record it is clear that Torsney suffers from neither a mental illness or defect nor a personality disorder which renders him dangerous to himself or others. He should, therefore, be released pursuant to the conditions properly imposed by the hearing court.8

Accordingly, the order of the Appellate Division should be reversed, with costs, and the order of Supreme Court, Kings County, reinstated.

(concurring). I concur in reversal and reinstatement of the order of Mr. Justice Yoswein directing the release of Torsney on condition. I agree that committing a person who has been acquitted under CPL 330.20 to the custody of the Commissioner of Mental Hygiene, without a prior hearing, is constitutionally permissible, but disagree that, except in the most formal sense, this results, as the plurality opinion suggests by its quotation from People ex rel. Henig v Commissioner of Mental Health (43 NY2d 334), from the fact that the individual "has successfully asserted an insanity defense”. The quoted words imply, and both the Henig opinion (43 NY2d, supra, at p 338) and that in People v Lally (19 NY2d 27, 33) flatly state, that what sustains the commitment is the individual’s success in convincing the trier of fact that he was insane when the crime was committed. To the contrary, as the plurality opinion recognizes (at p 674), but the dissenting opinion (at p 689) elides, the burden of proving a defendant’s sanity being on the State, acquittal by reason of mental disease or defect means nothing more than that the People failed to prove sanity, and therefore culpabil*685ity, beyond a reasonable doubt. Though there has, therefore, been no determination of "mental illness” sufficient to warrant commitment under the Mental Hygiene Law, commitment is, nevertheless, constitutionally permissible for "examination and report as to * * * sanity” (People v Lally, supra, at p 33; plurality opn, at p 674) in light of the prompt hearing required following such a commitment (People ex rel. Henig v Commissioner of Mental Health, supra, at p 340), of the need to protect society from the violent antisocial behavior which by its verdict of "not guilty by reason of mental disease or defect” the jury has necessarily found the individual to have committed (id., at p 338; see, also, Fhagen v Miller, 29 NY2d 348; and Matter of Coates, 9 NY2d 242), and of the individual’s act in entering such a plea.

CPL 330.20 does not, however, deal with how long an individual committed under its provisions may be held or what the relationship of his commitment to mental illness is. It authorizes a hearing at the instance of the individual (subd 5) or of the Commissioner of Mental Hygiene (subd 2), and conditions discharge (or release upon condition) only upon whether that may be done "without danger to himself or to others” (subds 2, 3; cf. subd 4). In my view, examination pursuant to temporary detention under CPL 330.20 must be followed either by release pursuant to hearing under CPL 330.20, if that be the commissioner’s and the court’s conclusion, or by a commitment hearing under the provisions of the Mental Hygiene Law, if the conclusion be that the detainee suffers from "mental illness”1 warranting warranting commitment under the Mental Hygiene Law, or both suffers from such an illness and is dangerous to himself or others. The point is, as the plurality decision concedes (at p 675), that dangerousness alone is an insufficient basis for involuntary commitment, and that while the commissioner or the detainee *686may show the detainee’s entitlement to release by showing his lack of dangerousness, the absence of such proof does not authorize continuation of involuntary commitment2 beyond a reasonable time for purpose of examination and report (cf. O’Connor v Donaldson, 422 US 563, 574-575; Jackson v Indiana, 406 US 715, 738; McNeil v Director, Patuxent Inst., 407 US 245, 249-250).

In summary, my view is that the only criterion for a CPL 330.20 release hearing is danger to self or to others, and that if the State intends to continue a CPL 330.20 commitment beyond the time required for examination and report it must begin a Mental Hygiene Law proceeding and make the showing required for any other civil commitment under that law. Absent such a showing the detainee will be entitled to release, on constitutional grounds via habeas corpus, even though he does not prove in a CPL 330.20 hearing by a preponderance of evidence that he is not dangerous (cf. Jackson v Indiana, 406 US 715, 736, 738, supra).

(dissenting). The court holds today that a man committed to a psychiatric institution because he shot and killed a young boy should be released despite expert psychiatric testimony that under similar circumstances he would be prone to act again with the same uncontrolled violence. I cannot agree with such a result and therefore dissent. The record offers insufficient proof that the appellant, Robert Torsney, no longer suffers from the very symptoms and personality disorder that precipitated the tragic killing of Randolph Evans. Releasing Torsney at this time would not only deprive him of the opportunity for rehabilitation offered in a psychiatric institution, but would also subject the public to the danger that Torsney might again be overcome by uncontrollably destructive impulses.

Torsney was charged with second degree murder for having fatally shot 15-year-old Randolph Evans on November 25, 1976. A New York City police officer on duty in Brooklyn, Torsney fired his weapon without provocation or justification at point blank range at Evans who was on his way home after walking his grandmother to a bus stop. At trial Torsney admitted having shot Evans, but contended that he was not criminally responsible for his actions when the shooting occur*687red. On November 30, 1977 a jury found him not guilty by reason of mental disease or defect (Penal Law, § 30.05).

After the trial Torsney was committed to the custody of the Commissioner of the Department of Mental Hygiene (CPL 330.20, subd 1). On July 20, 1978 the commissioner filed a petition in Supreme Court seeking Torsney’s release. Represented by independent counsel, Torsney also participated in the proceeding. After conducting a hearing Special Term ordered that he be released subject to certain conditions. The Appellate Division reversed, finding that Torsney had not sufficiently proven that he was no longer a danger to himself or others.

There are essentially two questions in this case: (1) what must a person committed pursuant to CPL 330.20 (subd 1) prove to secure his release? And (2) did Torsney meet the burden of proof?

The statute which prescribes the standard for release, CPL 330.20 (subd 3), provides: "If the court is satisfied that the committed person may be discharged or released on condition without danger to himself or others, the court must order his discharge, or his release on such conditions as the court determines to be necessary. If the court is not so satisfied, it must promptly order a hearing to determine whether such person may safely be discharged or released.”

Although the statute indicates that the critical factor for the court to consider on a petition for release is whether the committed person would pose a "danger to himself or others”, no guidance is provided as to what constitutes a "danger”. Certainly the term is open to various interpretations. In resolving the ambiguity, the Appellate Division construed the term to mean "likely to result in serious physical harm” (66 AD2d 281, 288) which in turn is defined as a "substantial risk of physical harm” (see Mental Hygiene Law, § 9.37, subd [a]). I agree, and the plurality at least seems to agree, that this is a proper and workable definition of dangerousness. However, the plurality would require the discharge of a detainee even though his release would "likely * * * result in serious physical harm to himself or others” if the dangerousness is not causally connected to an identifiable mental disease or defect. In addition the plurality states that to justify confinement of the detainee the mental disease must be treatable. I disagree.

I would hold that the detention of a person committed to a mental institution in accordance with CPL 330.20 (subd 1) *688should continue unless he proves that he no longer suffers from the symptomatology which made him dangerous.

It is well settled that the person committed pursuant to CPL 330.20 is presumed dangerous until he proves otherwise. As we so recently and unanimously held in Matter of Lublin v Central Islip Psychiatric Center (43 NY2d 341, 344): "Given the clear existence of this condition, as evidenced by the admitted commission of a violent act, it is appropriate that the condition be presumed to continue until the contrary is proven” (emphasis added; People ex rel. Henig v Commissioner of Mental Hygiene, 43 NY2d 334, 340; People v Lally, 19 NY2d 27). Despite this clear pronouncement, the plurality states "that the presumption flowing from an acquittal by reason of mental disease or defect does not presume that a person so acquitted presently suffers from mental disease or defect” (at p 674).

The sole issue confronted in Matter of Lublin v Central Islip Psychiatric Center (43 NY2d 341, 344, supra) and the resolution of that issue was stated as follows: "Must a person who has been validly committed as a result of an acquittal by reason of mental defect or disease, and who at a later time seeks release on the claimed ground that he is no longer dangerous to either himself or others, be required to prove by a fair preponderance of the evidence that he may safely be released? We hold that he must.” A person acquitted of a violent crime by reason of mental disease or defect has unequivocally demonstrated his dangerousness to society if not himself. Although involuntary confinement substantially deprives the committed person of liberty, a fundamental right, the State parens patriae interest in the mentally defective as well as its obligation to ensure public safety require on balance that the burden of proof fall to the detainee (Lublin, supra, at p 345).

There is still the question, however, of what the detainee must prove to secure his release. A resolution to this question depends in part on CPL 330.20 (subds 1, 2) which, in prescribing the standards and procedures for the discharge of detainees, makes no mention whatsoever of any requirement of mental illness to justify continued confinement. The sole consideration, according to the statute, is whether the person "may be discharged * * * without danger to himself or to others” (see McKinney’s 1960 Session Laws of NY, p 2024).

*689The verdict of not guilty by reason of mental disease or defect presumably resulted from a finding by the jury that the defendant at the time of the wrongful act suffered from symptoms which substantially limited his appreciation of "[t]he nature and consequence of such conduct” or "[t]hat such conduct was wrong” (Penal Law, § 30.05). The purpose for institutionalizing someone acquitted for this reason is twofold: to protect society from the threat that, because of a mental disorder, the detainee might again commit acts of violence (People v Lally, 19 NY2d 27, 33, supra), and to rehabilitate him so that he may be released into society without the risk of further harm. Should the symptoms of the mental disorder disappear both justifications for continued confinement disappear as well. No longer is the detainee, because of a mental disorder, a threat to the safety of himself or others. Nor once the symptoms have finally been remitted, is he in need of rehabilitation. Detention beyond this point would constitute punishment which is impermissible since the confined person was acquitted by the jury and found not to have been responsible for his actions when the wrongful act was committed. It is therefore logical that to secure his release the detainee should have to prove that he no longer suffers from the symptomatology associated with the wrongful act (Insanity Defense in New York: A Report to Governor Hugh L. Carey, New York State Department of Mental Hygiene, at p 50).

But I cannot agree with the plurality that a dangerous detainee must be released if he does not suffer from a "mental disease or defect” which fits into a psychiatric category and has a particular psychiatric label. Over the years the definition of mental illness has shifted according to the prevailing doctrine of psychological thought. In addition, there is often a lack of agreement among psychiatrists as to whether a given mode of unusual or deviant behavior constitutes a "mental illness” at all. We would therefore be unwise to bind our legal determinations to psychiatric theories which are not only undergoing constant re-examination and modification, but which are designed to deal with medical rather than legal problems. Our primary concern must be to protect the public from those who, for psychological reasons, have caused and are likely in the future to cause serious harm to others. To release such a person merely because his symptoms elude classification is indefensible. The issue must be resolved, not by the rigid application of psychiatric categories of mental *690illness, but rather by reasoned judgment based on careful scrutiny of the record.

Indeed by the very language of section 30.05 of the Penal Law the "mental disease or defect” defense embraces "mental defects”, a term of far broader reach than the less flexible standard of "mental disease”. It is quite possible that a jury could acquit a defendant by reason of a "mental disease or defect” when the defect was evidenced by symptoms not easily labeled as a "disease” listed in a psychiatric lexicon. Yet in the judgment of the jury the symptomatology might have prevented the defendant from appreciating the consequences of his conduct, thereby justifying application of the "mental disease or defect” defense. It would be a frightful paradox to release a detainee still manifesting the same symptoms because his mental defect could not be labeled in psychiatric terms — even though the jury might well have predicated its acquittal exclusively on the symptoms. By its holding today the plurality exalts this paradox to law.

To accomplish this dubious result, the plurality contends that it is following the constitutional mandate to afford detainees equal protection of the laws. Because a person cannot be civilly committed unless he is "mentally ill”, the plurality holds that a person who has been acquitted of a killing by reason of mental disease or defect must be released unless he too is "mentally ill”, despite the absence of any such requirement in CPL 330.20. The equation fails because there is a critical distinction between the two classes; the person acquitted of a violent crime by reason of mental defect has necessarily demonstrated by his own hand that he is a menace to society, whereas the person sought to be civilly committed has not. The State may not ignore the greater threat posed by the acquitted person. The compelling State interest to shield the public from this threat justifies that the two classes be dealt with somewhat differently (cf. People ex rel. Wayburn v Schupf, 39 NY2d 682; Matter of Richard E. R., 52 AD2d 927). Indeed the plurality itself recognizes that "the unique status of persons acquitted of a crime by reason of mental disease or defect permits the State to treat this 'exceptional class’ somewhat differently from other individuals believed suffering from mental disease or defect” (at p 673).

Nor can I agree with the plurality’s position that institutionalization may not continue if the mental defect is untreatable. Surely if a person acquitted of a violent crime by reason *691of a mental defect is compulsively homicidal because of a psychiatric disorder, the fact that the disorder is untreatable does not justify the person’s release and return to society.

Before turning to the hearing record on the application for release, it is instructive to review briefly reports of the psychiatrists and psychologists presented at the murder trial. Dr. Herbert Spiegel concluded that Torsney is "prone to hysterical dissociation under stress”, and Dr. Florence Shumer found Torsney to be "volatile” and "impulsive”. According to Dr. Daniel Schwartz, Torsney was suffering from psychomotor epilepsy. Dr. Joy Roy’s examination of Torsney revealed that "[w]hen he feels attacked and threatened the aggression escapes the censor and may erupt in impulsive behavior.” When this trial evidence is compared to the testimony presented at the hearing, it becomes all too apparent that Torsney’s condition has never changed substantially and that he remains as dangerous today as he was on the day he shot Randolph Evans.

At the hearing Dr. Stanley Portnow, one of the two independent court-appointed psychiatrists, noted that Torsney was not suffering from a classical form of psychosis or neurosis but rather labeled Torsney’s condition a "personality disorder”. Dr. Portnow further stated: "My diagnosis would have been constant throughout the entire spectrum of events and that diagnosis would read 'impulsive or explosive personality’ ”. "Stress”, according to Dr. Portnow, "is frequently the trigger by which the impulsive personality fires off”. He also expressed his belief that Torsney should face stress in small quantities to be reasonably sure that another catastrophe would not occur. Explaining why he recommended a gradual release procedure the doctor said: "I attempted to protect both Mr. Torsney and the community by recommending a prerelease procedure to minimize the severity of the stress upon this impulsive personality which already, on one occasion, has exploded, the result having been a catastrophe.” The other court-appointed psychiatrist, Dr. Milton Hollar, agreed that Torsney is "impulsive”, "volatile” and "explosive”. Dr. Hollar also feared that under stressful circumstances Torsney might erupt again. Nevertheless the doctor recommended Torsney’s release because as a nonpsychotic patient, Torsney’s institutionalization was not medically warranted.

During the period that Torsney was detained in the Mid-Hudson Psychiatric Center his principal therapist was Dr. *692Mark Vandenbergh who conducted five to eight sessions with Torsney of about one hour each. Dr. Vandenbergh stated that Torsney was susceptible to hysterical dissociative reactions under stress. It was such a reaction that precipitated the slaying of Evans, and according to Dr. Vandenbergh a violent explosion could occur again under similar circumstances. In view of these findings, Dr. Vandenbergh’s conclusion that Torsney was neither mentally ill nor in need of treatment are of little moment. Likewise, Dr. Alan Halpern, a staff psychologist at Mid-Hudson who examined Torsney only once, found some evidence of a neurotic character disorder — an hysterical personality. Dr. Halpern also observed that Torsney consistently seemed to lose his emotional control when faced with emotionally laden test situations. The only other therapist at the Mid-Hudson facility to examine Torsney was Dr. Easwara Bhoothalingom, the Forensic Unit Chief. Although he, too, examined Torsney but once, he nevertheless confirmed the familiar diagnosis of "hysterical reaction dissociative type”. It is noteworthy too that Dr. Bhoothalingom conceded that no special treatment had been administered to Torsney at Mid-Hudson.

Dr. Leslie Bowling, a staff psychologist at Creedmoor Psychiatric Center, was appointed by the court to administer a battery of psychiatric tests to Torsney, who had not previously been tested at Creedmoor, although he had already been there for six months. When asked if Torsney, because of his impulsive personality, had the capacity to act the same way as he had acted on the day of the killing, Dr. Bowling answered "yes”, but added that there was no way to be certain. Nevertheless because he deemed Torsney’s condition untreatable, Dr. Bowling recommended release despite the apparent danger.

Insofar as the findings of the special release committee, comprised of Drs. Sablón Dartigue, Joseph Sklar and John McKnight, are concerned, it is clear that only Dr. McKnight had had any significant contact with the patient. Indeed Dr. Sklar admitted that he was completely unfamiliar with Torsney’s background. Dr. McKnight diagnosed Torsney’s condition as "hysterical neurosis dissociative type in remission”, marked by inappropriate reaction to stress. Yet he recommended that Torsney be released on condition that he not be a policeman. Similarly the independent review panel at Creed-moor conducted no independent psychological evaluation of *693Torsney, but instead based its recommendation on a perusal of Torsney’s hospital records and an hour and a half interview. Interestingly, Dr. Louis McGarry, a member of the panel, stated that Torsney’s personality disorder was treatable, but that the patient was not receptive to treatment.

It is most important to remember that in analyzing this testimony, the trier of the facts need not accept the ultimate conclusions of expert witnesses regarding a patient’s dangerousness or the propriety of his continued confinement, for the underlying facts as attested to by the witnesses may support contrary conclusions (cf. Cross v Harris, 418 F2d 1095, 1100-1101). Hence although several psychiatrists testified that Torsney should be released, I would agree with the view expressed by the Appellate Division that their testimony strongly indicates the need for his further confinement.

The hearing testimony overwhelmingly demonstrates that Torsney has remained as volatile and explosive as he was on the date of the killing. The evidence shows that his explosiveness is a manifestation of a personality disorder as an hysterical dissociative reaction, the same diagnosis offered by Dr. Spiegel at the murder trial. According to these experts it was this disorder that impelled Torsney to kill Randolph Evans, and according to the same experts it is this disorder that could drive Torsney to commit a similar act of violence in the future. Clearly Torsney’s condition has not changed. Indeed he has not received individualized treatment at either Mid-Hudson or Creedmoor.

The plurality is concerned that we not " 'transform the hospital into a penitentiary where one could be held indefinitely’ ” (at p 677, quoting Ragsdale v Overholser, 281 F2d 943, 949-950). I agree, of course, that hospital confinement should never be abused to deprive someone of his liberty unfairly. However, when a person, driven by an explosive personality disorder, has proven by an act of senseless violence that he is a menace to others, so long as the psychological condition persists it is our obligation to protect the public from further acts of violence that he may commit.

Accordingly, I would affirm the order of the Appellate Division.

Judges Jones and Fuchsberg concur with Judge Jasen; Judge Meyer concurs in a separate opinion in which Judge Fuchsberg also concurs in the following memorandum: Since *694I see no inconsistency between the views expressed by Judge Jasen in the plurality opinion and those by Judge Meyer in his concurrence, I join in both; Judge Wachtler dissents and votes to affirm in another opinion in which Chief Judge Cooke and Judge Gabrielli concur.

Order reversed, etc.