MEMORANDUM *

Jennie Catalano appeals the district court’s decision affirming the Administrative Law Judge’s (“ALJ”) denial of Social Security disability benefits. Because the ALJ offered clear and convincing reasons for partially discrediting Catalano’s testimony, and because substantial evidence supports the ALJ’s findings, we affirm.

We review de novo the district court’s decision upholding the ALJ’s denial of Social Security disability benefits. See Tommasetti v. Astrue, 533 F.3d 1035, 1038 (9th Cir. 2008). The ALJ’s decision must be affirmed if it applied the correct legal standard and is supported by substantial evidence. See Parra v. Astrue, 481 F.3d 742, 746 (9th Cir. 2007).

I. Credibility of Catalano’s Testimony

“To determine whether a claimant’s testimony regarding subjective pain or symptoms is credible, an ALJ must engage in a two-step analysis.” Lingenfelter v. Astrue, 504 F.3d 1028, 1035-36 (9th Cir. 2007). “First, the ALJ must determine whether the claimant has presented objective medical evidence of an underlying impairment ‘which could reasonably be expected to produce the pain or other symptoms alleged.’ ” Id. at 1036 (quoting Bunnell v. Sullivan, 947 F.2d 341, 344 (9th Cir. 1991) (en banc)). Catalano satisfied this step by presenting objective medical evidence of chest pains and physical deconditioning. Second, if “there is no affirmative evidence of malingering, ‘the ALJ can reject the claimant’s testimony about the severity of her symptoms only by offering specific, clear and convincing reasons for doing so.’ ” Tommasetti, 533 F.3d at 1039 (quoting Smolen v. Chater, 80 F.3d 1273, 1281 (9th Cir. 1996)). Because there was no affirmative evidence of malingering, clear and convincing reasons were required to discredit Catalano’s testimony.

To the extent the ALJ in fact discredited Catalano’s testimony, he provided *603three valid reasons for doing so. First, the ALJ found Catalano was unable to testify with specificity regarding her symptoms. See Tommasetti, 538 F.3d at 1040 (ALJ may “rely on ordinary techniques of credibility evaluation,” such as whether claimant was a “vague witness”). Second, the ALJ evaluated Catalano’s testimony in light of the objective medical evidence in the record. See Lingenfelter, 504 F.3d at 1040 (ALJ may consider “whether the alleged symptoms are consistent with the medical evidence” when assessing credibility). Third, the ALJ observed Catalano’s appearance and demeanor at the hearing, at which she did not have any apparent limitations. See Orn v. Astrue, 495 F.3d 625, 639 (9th Cir. 2007) (ALJ’s “observations of a claimant’s functioning” at the hearing are permissible as part of the overall credibility assessment). Though perhaps none of these reasons by itself would be clear and convincing, collectively they provide a sufficient basis for partially discrediting Catalano’s testimony. See Tommasetti, 533 F.3d at 1039 (upholding adverse credibility determination where the “ALJ provided several permissible reasons”).

II. Catalano’s Disability Claim

A claimant is disabled only if she has a severe impairment, or combination of impairments, significantly limiting her physical or mental ability to do basic work activities. 20 C.F.R. §§ 404.1520(c), 1521(a). Substantial evidence supports the ALJ’s finding that Catalano is not disabled because her residual functional capacity (“RFC”) enables her to perform her past relevant work.

As a preliminary matter, the record does not support Catalano’s claim that the ALJ failed to consider a psychological cause for her symptoms: the ALJ noted the possibility that Catalano’s symptoms were psychosomatic, and Dr. West opined that they may be caused by aggravation or emotional factors. Catalano’s mental impairment claim also fails because there is no evidence of a functional limitation related to daily activities, social functioning, or concentration. See 20 C.F.R. § 404.1520a(e)(3). On the contrary, psychiatrist Dr. Carone reported no problems with Catalano’s intellect, judgment, manner of relating to others, or ability to engage in daily activities.

Catalano’s challenges to the ALJ’s RFC determination fail because the ALJ considered her capacity for sustained work, and Catalano never established any non-exertional impairments. RFC is the most a claimant can do in a work setting on a regular basis, despite her limitations. See 20 C.F.R. §§ 404.1545(a), (b), (c). After noting that employment would be difficult for someone whose impairments caused frequent absences from work, Dr. West concluded Catalano remained capable of sedentary to light work. To the extent Dr. West’s testimony may have been ambiguous, this court will not second-guess the ALJ’s reasonable interpretation. See Edlund v. Massanari, 253 F.3d 1152, 1156 (9th Cir. 2001) (“The ALJ is responsible for determining credibility, resolving conflicts in medical testimony, and resolving ambiguities,” and where “the evidence is susceptible to more than one rational interpretation, the court may not substitute its judgment for that of the [ALJ]”). Notably, Dr. West’s assessment was not inconsistent with Catalano’s own testimony, and it was in accord with the reports of every other treating, examining, and non-examining physician in the record, none of which indicated Catalano was incapable of sedentary to light work. Substantial evidence thus supported the ALJ’s “sedentary to light” RFC determination.

*604When applied to her past relevant work, this RFC renders Catalano not disabled. Catalano does not allege her past relevant work involved more than “light” exertion, and the vocational expert testified that it is performed at the “sedentary” level in the national economy. A claimant is not disabled if she is capable of performing her past relevant work. 20 C.F.R. § 404.1520(f). Therefore, the ALJ properly concluded that Catalano was not disabled because her “sedentary to light” RFC enables her to perform her past relevant work.

AFFIRMED.

dissenting.

Often in this circuit, when we decide a case by memorandum disposition, we include language stating that “because the parties are familiar with the facts, we do not recite them.”1 This language has always struck me as odd, since the parties are “familiar with the facts” in every case. This case, however, shows why reciting at least some facts can be important: It prevents us from basing our decisions on illusory records and arriving at decisions that would be inconceivable under the actual records before us.

I am not sure to what record my colleagues are referring when they say that “the ALJ offered clear and convincing reasons for partially discrediting Catalano’s testimony” that her medical impairment caused her to miss work regularly. Maj. op. at 602. They are most certainly not referring to the record in this case. In this case the ALJ did not even identify which portions of Catalano’s testimony he was rejecting, nor did he offer any reasons for doing so, let alone provide specific, clear and convincing reasons for any such rejection. Moreover, the actual record reveals that there was no testimony or other evidence that contradicts or undermines the part of Catalano’s testimony that the majority appears to speculate that the ALJ rejected, or indeed any part of Catalano’s testimony. All in all, it is difficult to conceive of an ALJ decision that could violate more fundamental rules in one short opinion.

The majority and I do not disagree with regard to the law. We all recognize that because “pain and other such symptoms” are “highly subjective and idiosyncratic,” Smolen v. Chater, 80 F.3d 1273, 1282 (9th Cir. 1996), a claimant’s own “testimony as to subjective symptoms” cannot be discredited “merely because [it is] unsupported by objective evidence.” Lester v. Chater, 81 F.3d 821, 834 (9th Cir. 1996). To the contrary, as the majority purports to understand, “the ALJ can reject the claimant’s testimony about the severity of her symptoms only by offering specific, clear and convincing reasons for doing so.” Maj. op. at 602 (emphasis added) (internal quotation marks omitted) (quoting Tommasetti v. Astrue, 533 F.3d 1035, 1039 (9th Cir. 2008)). My colleagues also undoubtedly recognize that “[t]he clear and convincing standard is the most demanding required in Social Security cases,” Moore v. Comm’r of the Social Security Admin., 278 F.3d 920, 924 (9th Cir. 2002), and that to satisfy this demanding standard, the ALJ must make “specific findings [,] supported by the record,” explaining why he believes the claimant is not credible with respect to some or all of her testimony. *605Bunnell v. Sullivan, 947 F.2d 341, 346 (9th Cir. 1991) (en banc) (emphasis added).

Given these agreed upon legal rules, it is incomprehensible how the majority could come to the conclusion it does in this case. It could do so only by ignoring the facts and the actual record, and unthinkingly regurgitating abstract legal principles instead of undertaking the hard work judges are supposed to perform when applying the law to the facts. It is unfortunate for Mrs. Catalano that after waiting over twenty years to receive the benefits to which she is entitled she fell victim to such regrettable judicial treatment.

Our analysis must begin by seeking to determine precisely what “testimony” the ALJ “partially discredited].” Maj. op. at 602. Here we run into our first encounter with the majority’s failure to base its decision on the facts in the record: Contrary to what one might think from reading the majority’s brief disposition, the ALJ in this case did not say what parts, if any, of Catalano’s testimony he was rejecting. To the contrary, he held that Catalano’s “testimony regarding [her] pain and other symptoms [was] credible and reliable to the extent it [was] consistent with [his] decision.” How the majority sees in this statement a finding that a specific portion of Catalano’s testimony was not credible is wholly beyond me. At best, all that the ALJ has provided us is a vague statement that he discounted some part of her testimony. That is plainly insufficient. The ALJ’s “findings, properly supported by the record, must be sufficiently specific to allow a reviewing court to conclude the adjudicator rejected the claimant’s testimony on permissible grounds.” Id. at 345 (emphasis added). We cannot properly review the ALJ’s rejection of Catalano’s testimony if we do not know what testimony he rejected.

The majority commits the same offense as the ALJ. It leaves us to guess what portions of Catalano’s testimony it guessed the ALJ rejected. Athough it is far from clear, the most plausible guess that a reader might make as to the testimony the panel had in mind would be Catalano’s statements regarding the frequency of the episodes that rendered her unable to hold a regular job. If she were unable to work for two or three days each time she suffered from one of her regular episodes, then, as Dr. West — the medical expert upon whom the ALJ relied — testified, she would “have a hard time maintaining employment.”

Let us be clear as to what Catalano’s testimony actually was: She stated that her psycho-physiological condition, which the ALJ specifically and unequivocally found to be a “severe impairment,” caused her to “crack[] many times under pressure and landed [her] in the emergency room very often, having to stay home each time this happened.” In a letter admitted into the administrative record that Catalano wrote in 1981, when she was actually suffering from these episodes, she stated that her nervous attacks involved “deadening chest pains ... dizziness, nausea, diarrhea, plus numerous other symptoms [that rendered her] .... completely useless for at least two days ” after each of the numerous attacks. The majority apparently believes that the ALJ discredited these statements because he ultimately found, despite Catalano’s missing work on a regular basis for days at a time, that her “residual functional capacity enable[d] her to perform her” prior job. Maj. op. at 603. Curiously, the ALJ specifically acknowledged that “[d]ue to her missing so many days at work, [Catalano’s] files were taken away and given to others in the office.” Given that the ALJ acknowledged that Catalano had to “miss[ ] so many days at work,” it is impossible to determine pre*606cisely what caused him to conclude that she had the residual functional capacity to perform her job, and it is not at all clear that he discredited Catalano’s statements that her condition caused her to miss work regularly for days at a time. The ALJ’s decision is absolutely silent on this point, which is why it was necessary for the majority to attempt to guess at what part of her testimony, if any, he actually discredited, and why it is particularly unfortunate that the majority fads to enlighten us as to what its guess was.

Although the ALJ’s failure to state what portions, if any, of Catalano’s testimony he disregarded is reason enough to reverse the decision below, the majority’s creative reading of the record does not stop there. After deciding that the ALJ must have rejected some particular part of the testimony but failing to tell us what part it was, the majority next provides on behalf of the ALJ his reasons for rejecting that unspecified testimony, even though those reasons are neither stated in nor supported by the record. Because the ALJ failed to state that he was rejecting Catalano’s testimony for any specific reasons, the majority has invented, post hoc, three explanations that the ALJ might have offered for his decision. It does so by seizing on three random statements at various points in the ALJ’s opinion, at least two of which were not offered by the ALJ as a reason for his credibility finding. In so doing, the majority engages in the very “speculation] as to the grounds for an adjudicator’s rejection of a claimant’s allegations of disabling pain” that our case law prohibits. Bunnell, 947 F.2d at 346.

What is more, the explanations that the majority puts forward on the ALJ’s behalf are themselves either arbitrary and capricious or wholly irreconcilable with the record. The majority’s first contention is that the ALJ discredited Catalano’s testimony because he found her to be a “vague witness.” Maj. op. at 603. However, as the actual record reflects, the ALJ never made such a finding. The ALJ observed that “Catalano could not testify with any specificity as to her daily symptoms.” This is neither remarkable — since, as the ALJ stated, the symptoms occurred “twenty years” prior to her testimony — nor relevant. Regardless of the specificity of her testimony as to what happened on a daily basis, the ALJ found that Catalano’s medical condition did constitute a “severe impairment.” The relevant question is whether she suffered from regular episodes resulting from that condition frequently enough to cause her to miss substantial periods of work. On this question, the ALJ stated, as quoted above, that “[d]ue to her missing so many days of work, [Catalano’s] files were taken away and given to others in the office.” He also related the testimony of Dr. West that the hospital records showed multiple admissions for complaints of chest pains during the relevant period. Far from finding that Catalano’s testimony was “vague” with respect to whether her “severe impairment” caused her to miss work regularly, the ALJ implicitly acknowledged that she did so.

The majority next suggests, on the ALJ’s behalf, that the ALJ rejected Catalano’s testimony because it was in conflict with the “objective medical evidence in the record.” Maj. op. at 603. The sole evidence to which the majority refers is the medical expert’s “conclu[sion that] Catalano remained capable of sedentary to light work.” Id. at 603. However, a reading of the actual record demonstrates that the majority takes this statement entirely out of context and that read properly the statement in no way conflicts with Catalano’s testimony. Dr. West’s testimony was that Catalano’s physical deconditioning— meaning the fact that she was, in general, *607less physically fit “than the average 55-year-old” woman — would place her in the “sedentary to light” work-capacity range. He specifically said that he made this statement “because [Catalano] is deconditioned.” He was not speaking about her impairment or its consequences when he stated that her “deconditioned” state rendered her capable of only “light to sedentary” work. When he was asked later about the impact of her impairment on her residual functional capacity, he testified that “if somebody perceives that they have a severe ... chest ache or an abdominal pain ... if those symptoms are sufficiently debilitating, [that person] will have a hard time maintaining employment.” Although he did not know the cause of Catalano’s impairment, which he said could have resulted from “emotional factors, could be esophegeal, could be other GI etiology, could be pulmonary, [or] could be cardiac,” he specifically said that “employment is going to be difficult for a lady who, for whatever reason, is going to the emergency room quite frequently because of perceived chest pain.” He later said, referring specifically to Catalano, that “a major problem is going to be absence from work.” Dr. West’s testimony that frequent hospital visits will severely interfere with someone’s employment capacity applies to any individual, whether ordinarily her work classification would be “sedentary to light,” medium, heavy, or otherwise. If, as the majority assumes, the ALJ relied on Dr. West’s testimony to reject Catalano’s statements regarding the frequency and effect of the heart pains that regularly landed her in the hospital and caused her to miss work for two or three days at a time, then he did so without clear and convincing reasons supported by the record, because the doctor’s testimony is not at all inconsistent with those statements.2 At no point did the medical expert suggest that the episodes Catalano experienced were nonexistent, exaggerated or, most important, inconsistent with her underlying impairment.

The majority’s third explanation for why the ALJ rejected part of Catalano’s testimony is that “the ALJ observed Catalano’s appearance and demeanor at the hearing, at which she did not have any apparent limitations.” Maj. op. at 603. This is the most preposterous of the justifications the majority offers on the ALJ’s behalf. Catalano testified before the ALJ twenty years after her “severe” and debilitating attacks caused her to miss work on a regular basis. What possible relevance is there in the fact that the long-since retired Catalano, testifying in 2002, “did not have any apparent limitations”? The question at issue is whether her traumatic episodes in 1980 — when she worked full-time in a position that the ALJ acknowledged was “of a high stress nature” — occurred so frequently that she had to miss substantial periods of work. Catalano’s uncontroverted testimony is that they did. Her physical appearance at a hearing twenty years later, following a long period of retirement, has absolutely no bearing on that question, and if the ALJ did in fact rely on this ground, then any decision he made to reject Catalano’s testimony was arbitrary and capricious and made without clear and convincing reasons supported by the evidence, and must be rejected. Cf. Orn v. Astrue, 495 F.3d 625, 639 (9th Cir. 2007) (“[An] ALJ’s observations of a claimant’s functioning [at the hearing] may not form the sole basis for discrediting a person’s testimony.”).

*608In sum, the question the ALJ was required to determine was whether, in light of the “severe impairment” he found Catalano to have had, she had a sufficient residual functional capacity to perform her job as an office manager and insurance clerk. The key factual question in this inquiry was how regularly her impairment caused her to miss work. As the government’s own medical expert acknowledged, “a major problem [for someone with Catalano’s condition] is going to be absence from work,” and no one disputes that absence from work for two or three days at a time on a regular basis renders a person unable to hold a job. There is no evidence in the record, medical or otherwise, to contradict Catalano’s testimony that her impairment in fact “landed [her] in the emergency room very often, having to stay home each time.” Nor did any doctor or medical expert suggest that the subjective responses she experienced were in any respect inconsistent with her underlying impairment. Presumably the ALJ reached a contrary conclusion on his own, a conclusion that was in conflict with all the evidence in the record.3

For no reason evident in the record, the ALJ concluded that Catalano’s “impairment [did not] preelude[ ] the performance of her past relevant work.” As to his actual reasons, if any, the record is silent. Under our clearly established law, the ALJ’s unexplained reasoning is insufficient to support his unexplained findings and unwarranted ruling. We cannot affirm a determination that is neither expressed nor supported by specific findings as well as by clear and convincing reasons. Nor can we affirm a determination that is contrary to all the evidence in the record. The ALJ’s ruling fails to meet our legal standards in every respect. I therefore vote to reverse.

For these reasons, I dissent.