OPINION

Jeffrey Benton, on behalf of his deceased wife Lynn Benton (“Benton”), appeals the district court’s affirmance of the Commissioner’s denial of her claim for Social Security Disability Insurance Benefits. An Administrative Law Judge (“ALJ”) found that Benton was not mentally disabled. In doing so, he refused to credit the opinion of Benton’s psychiatrist as her treating physician. The ALJ instead relied upon the opinion of a psychiatrist who examined Benton once at the behest of the state of California. According to a testifying vocational expert, the absence or presence of mental disability was the difference, between Benton’s being able to perform her past relevant work or not. Because we conclude that the ALJ misapplied 20 C.F.R. § 404.1502 , the regulation that addresses the definition of “treating source,” we reverse the district court, vacate the ruling of the Commissioner and remand.

BACKGROUND

A. PROCEDURAL TIMELINE

Benton pursued her claim for disability through the administrative process, appealing the denial to the district court on December 20, 2000. Benton died from the effects of a nasopharyngeal carcinoma on July 25, 2001; this condition was unrelated to her disability claims. Her husband has continued to pursue her appeal on her behalf, timely appealing denial of her claim.

B. FACTS

Benton was born on July 16, 1953. She obtained a high school education. She injured her leg while performing high school gymnastics in 1971. As a result, she required psychological treatment and underwent the first of seven knee surgeries; the others were performed annually from 1983 to 1987 and again in 1996. Subsequent to her accident she worked as a waitress, a receptionist, a telephone customer service representative, and an administrative assistant in a hotel.

1.Vocational history and level of functioning at daily activities

Benton stopped work on August 4, 1995 due to post-traumatic arthrosis in her left knee and pain in her back, hips, and feet. In the Disability Report she filed when she applied for benefits, she reported being unable to sit, stand, or walk for extended periods. She used a cane to walk; the Claims Representative noted her grimacing and needing to stand up occasionally. A later Disability Report indicated increasing difficulty entering and leaving the bathtub and shower, fixing her hair, writing, and engaging in handcrafts. In a Daily Activities Questionnaire and later testimony at a hearing, she reported depression and sleep deprivation due to chronic pain, and decrements in memory and concentration. She shopped using a mobility cart or a wheelchair pushed by her husband. Benton did do light housework during this period: she cooked one meal three to five days of the week, washed dishes, made the bed, dusted, and did laundry with her husband’s help. But she found this work laborious; she reported having to sit down for ten minutes after three to five minutes of washing dishes.

Regarding her ability to work, Benton reported that if she sat for more than 45 minutes, she would have spasms and her left foot would go numb. After ten to twenty minutes of sitting in an office chair, she would have to get up and move for five to ten minutes before sitting down. She could only stand in one place for three to five minutes, and could walk only about a half block with her cane. She had to elevate her left knee for relief, could not carry more than eight pounds, and lost her balance easily.

2. Medical history — physical

Benton was a patient of Kaiser-Perma-nente (“Kaiser”), a nonprofit, group-practice health maintenance organization (“HMO”). Benton was referred to orthopedic surgeon Dr. Donald Fithian in May 1991. He recommended delaying a total knee replacement for ten years, given her young age, in the hope that the combination of cortisone injections and abstaining from work would prove satisfactory until then. But the injections did not prove efficacious for long, and Benton’s x-rays showed progressive degenerative arthritis. She began feeling pain in her left shoulder, low back and hip, which Dr. Fithian hypothesized was due to her limp. On July 26, 1995, Dr. Fithian opined that she was permanently disabled due to pain in her lower left extremity. Her seventh surgery led to initial improvement in her knee pain, but not her hip. After she experienced a fall in April 1997, the pain management therapy became less effective, affecting her mood. At this time, Dr. Fithian recommended a knee replacement.

Other doctors employed by the State of California Health and Welfare agency and the Social Security Administration acknowledged the impairment of her knee and restrictions that it imposed but disagreed that it was disabling.

3. Medical history — psychological

In November 1997, Benton saw Dr. Zwiefach, a psychiatrist with Kaiser, who diagnosed Chronic Pain and Dysthemia. In treating her, he prescribed first Ela-vil and then Paxil, increasing the dose when she reported no effect. Dr. Zwie-fach continued managing her psychiatric medications and consulted regularly with her treating therapists. When asked to evaluate Benton for mental disorders, he observed that she had difficulty with concentration and reported psychomotor retardation, feelings of guilt and worthlessness. He diagnosed her with major depression, an unspecified personality disorder, and chronic pain. He assigned her a Global Assessment of Functioning rating of 53, representing moderate symptoms or moderate difficulty in, inter alia, occupational functioning.

At the request of the state agency, Benton was examined in September 1998 by Dr. Engelhorn, a psychiatrist who noted functional disability due to her chronic left knee pain. He noted onset of depression in late 1997. He reported that Benton was “fully capable of taking care of .her basic personal needs” and was “involved in a full variety of light household chores.” He found no cognitive impairment or evidence of significant depression or anxiety. He diagnosed “perhaps a mild adjustment type of reaction with low levels of depression perhaps beginning at the end of 1997,” which “appears to be an adjustment type of depression relating to her physical disabilities and great pain ... [that] totally relates to problems pertaining to her left knee.” State agency psychiatrist Dr. Sko-pec reviewed the medical records and opined that Benton had no severe mental impairment.

In January 1999, Dr. Zwiefach completed a Mental Residual Functional Capacity Assessment (“Mental RFC Assessment”) on Benton, finding marked limitations in 11 of 20 categories. He assigned her a prognosis of very poor.

4. Benton’s hearing before the ALJ

At her hearing before the ALJ, Benton testified that Dr. Fithian was her orthopedist and Dr. Zwiefach was the psychiatrist overseeing her case, although she had met with him only once. She testified that her right hip, left knee, and both feet bothered her most, and that she could sit comfortably for 10-20 minutes, stand for 5-10 minutes, and walk one-half block. She normally used a cane and still did light housekeeping, cooking, and crafts other than quilting.

The ALJ applied the five-step sequential process presented in 20 C.F.R. §§ 404.1520 (a)-(f) that is used to determine whether a claimant is disabled under the Social Security Act. This process is presented in full elsewhere, e.g. Tackett v. Apfel, 180 F.3d 1094, 1098-99 . We summarize it as follows: if a claimant cannot meet the burden of showing she is currently not working, at step one, the process ends; if she establishes that she meets the “severely impaired” criteria, at step two, she is “disabled” and entitled to benefits. However, when she has not met the step two criteria, if she can show at step three that her impairment satisfies certain specific criteria listed in the regulations, she is “disabled.” If she can show at step four that she is unable to perform work she has done in the past, she is entitled to a step five review. At step five, the government has the burden of showing that she can do other work available in significant numbers in the national economy; if it does not meet this burden, the claimant is deemed “disabled.”

The ALJ found that Benton met step one: she had not engaged in substantial gainful activity since August 4, 1995; and that although she had severe left knee degenerative joint disease and arthritis, chronic pain, and an affective disorder, she did not meet step two. A vocational expert testified that a hypothetical person with Benton’s physical residual function capacity could perform her past work as a receptionist, but could not sustain any competitive employment based upon either the mental assessment of Dr. Zwiefach or Benton’s testimony at the hearing as to her physical condition. The ALJ did not credit Dr. Zwiefach’s opinion as a treating doctor, and found Benton’s claims of disabling pain and limitations not credible. He found that Benton was able to perform sedentary work, and denied her benefits at step four of the sequential evaluation process, finding that she could return to her past relevant work as a receptionist.

DISCUSSION

A. Jurisdiction

The district court had subject matter jurisdiction to review the final decision of the Commissioner of Social Security pursuant to 42 U.S.C. § 405 (g). We have jurisdiction to review the district court’s final decision under 28 U.S.C. § 1291 .

B. Standard of Review

The district court decision affirming the ALJ is reviewed de novo. The Commissioner’s denial of disability benefits may be set aside only when the ALJ’s findings are based on legal error or not supported by substantial evidence in the record. If the evidence can support either outcome, the Commissioner’s decision must be upheld. Flaten v. Sec’y of Health & Human Servs., 44 F.3d 1453, 1457 (9th Cir.1995).

C. Issues regarding Dr. Zwiefach

Benton disputes the ALJ’s rejection of Dr. Zwiefaeh’s Mental RFC Assessment.

1. Does Dr. Zwiefach qualify as Benton’s treating physician?

The ALJ found that there is no evidence that Dr. Zwiefach qualifies as Benton’s treating physician. The applicable regulation, 20 C.F.R. § 404.1502 (2002), defines a “treating source” as follows:

The wording of the definition of “treating source” has remained constant since 1991; the provision of medical treatment, particularly for psychological dysfunction and particularly within HMOs, has not. HMOs achieve cost savings in part by shifting the provision of services to less costly providers, e.g. psychologists and counselors who provide services under the supervision of a psychiatrist. This is notably true in the domain of psychological dysfunction because, while provision of drugs such as the anti-depressants Benton took has increased substantially, in California only physicians have prescription privileges. A psychiatrist may therefore manage the provision of psychiatric medication, receiving reports from the medical sources providing “hands-on” treatment, without seeing the patient with any regularity. This was precisely Dr. Zwiefaeh’s role in Benton’s treatment.

Section 404.1502 neither explicitly forbids nor requires crediting a physician “treating” status whose patient contact is thus limited. Its language suggests that “a few times” or contact as little as twice a year would suffice, but it does not state that this frequency of patient contact represents a floor. Rather, the standard it applies is that the claimant must have seen “the source with a frequency consistent with accepted medical practice for the type of treatment and/or evaluation required for your medical condition(s).”

The opinions of treating physicians are given greater weight than those of examining but non-treating physicians or physicians who only review the record.

Lester v. Chater, 81 F.3d 821, 830 (9th Cir.1995) (citations omitted).

The Commissioner argues that the rationale for according greater weight to a treating doctor’s opinion depends on the existence of an ongoing treating relationship, which affords a greater opportunity to know and observe the patient, and that that opportunity was not present here. See Smolen v. Chater, 80 F.3d 1273, 1285 (9th Cir.1996). Benton argues that Dr. Zwiefach reported her symptoms as the leader of a treatment team that had an extensive ongoing treating relationship with her.

Obviously, one or more other members of that treatment team other than Dr. Zwiefach had sufficient contact with Benton to qualify unequivocally as a treating source. Dr. Zwiefach completed the mental assessment of Benton based on his assessment of information provided by those on the treatment team with more direct patient contact; the qualifications of those providing direct treatment, and the quality of their communications with Dr. Zwiefach, were either already present in the record or readily ascertainable by the ALJ. The question, then, is whether § 404.1502 precludes the supervising physician from being considered a treating source in such circumstances.

This is an issue of first impression for our court. Outside of our circuit the meaning of “treating source” in 20 C.F.R. § 404.1502 has been construed at the federal appellate level in only one published opinion. An ALJ found that a nurse practitioner, Ms. Flaherty, who filled out an RFC assessment was not to be given “treating source” weight:

Shontos v. Barnhart, 322 F.3d 532 , 539 (8th Cir.2003). The Eighth Circuit accepted the ALJ’s finding that Nurse Flaherty was not an “acceptable medical source,” as nurse practitioners are not listed in 20 CFR § 404.1513 (a), but found that she qualified as an “other medical source” under § 404.1513(d) and that her opinion could also be counted as a “medical opinion” per 20 C.F.R. § 404.1527 .

The Shontos court does not address whether seeing a patient one time was not enough. It does, however, find that the use of a team approach by medical providers is analytically significant:

Id. at 340 .

Although it does not address § 404.1502, the district court and the Commissioner in our case cite to Matney on Behalf of Matney v. Sullivan, 981 F.2d 1016 (9th Cir.1992) for the precedent that someone who has only seen a patient once is not a treating physician. As a redesignated (but not rewritten) unpublished opinion, Matney is sparing with the facts. The sentence of interest is “The ALJ determined that the opinion of Dr. Cookson was entitled to little weight because he examined Mr. Matney only one time and produced a brief report.” Id. at 1020 . Other than noting Matney’s claim that Dr. Cookson was his treating physician, the opinion gives no indication of the doctor’s specialty, if any, and what role he played in Matney’s treatment.

Matney is distinguishable from the instant case in numerous ways. First, although Dr. Zwiefach had seen Benton only once, he continued to oversee her care. Beyond that, Cookson provided only a “brief report,” his “clinical evaluation revealed very minimal abnormal findings” and he “agreed to become an advocate” in presenting Matney’s petition to the ALJ. Even if Matney were not otherwise distinguishable, the single sentence cited is not a finding that having seen Matney only once disqualified Dr. Cookson independent of these other factors. Matney also did not reflect any examination of whether Dr. Cookson’s frequency of treatment reflected “acceptable medical practice” at the time as per § 404.1502. Even if it had, such a finding would now require reexamination, because what constitutes acceptable practice has changed over the course of the intervening decade.

This circuit’s district courts have produced three decisions relevant to the question of who is a treating physician. Bowman v. Comm’r, Soc. Sec. Admin., 2001 WL 215790 (Feb. 23, 2001) 2001 U.S. Dist. LEXIS 4391 , credited a doctor who provided services after the expiration of a claimant’s insured status as being a treating physician for the purposes of the regulations. It states that “[t]he key issue in determining whether Dr. Gordon was claimant’s treating physician is whether his examinations of her were prompted by her need for treatment.” Id. at *12 . Joves v. Callahan, 1997 WL 362040 (June 24, 1997) 1997 U.S. Dist. LEXIS 9078 at *8 n. 1, states merely that a “treating source” is a “physician or psychologist” who has had an “on-going treatment relationship,” without elaborating on what “treatment” need entail.

The most helpful statement of the law regarding § 404.1502 comes from Magistrate Judge Jelderks in Ratto v. Sec’y, Dep’t of Health & Human Servs., 839 F.Supp. 1415, 1425 (D.Or.1993):

(citations to record omitted.) The court in Ratto found that Dr. Stites’s testimony should have been credited by any standard.

Here, Dr. Zwiefach had examined Benton not much more than a year before his report, and was still employed to cure her. Ratto would suggest that, as an individual, while Dr. Zwiefach may be placed relatively low on the continuum of treating physicians in this respect, he would still fall into the treating physician category. His opinion would be entitled to greater weight than that of an examining or reviewing physician.

This, however, understates the weight that Dr. Zwiefach’s opinion may be due. Dr. Zwiefach is transmitting both his own knowledge and opinion of Benton and those of the medical treatment team under his supervision. Even if he were not entitled to complete the Mental RFC Assessment based upon his own direct experience with a patient, nothing in the language of § 404.1502 forecloses his doing so on behalf of his treatment team. This is not the same as evaluating Benton’s case from the cold record; Dr. Zwiefach has had the opportunity to direct and communicate with the treatment team over time, and is presumably well placed to know their skills, abilities, and therapeutic techniques. In assigning weight to Dr. Zwiefach’s opinion, the ALJ may of course consider how well the treatment team operated in informing Dr. Zwiefach.

The ALJ did not consider whether Dr. Zwiefach saw Benton with a frequency consistent with accepted medical practice for this type of treatment, nor did he consider the evidence of Dr. Zwiefach’s ongoing prescription and medication management of Benton’s psychiatric evaluations and regular consultations with her therapists.

2. Did corroborating objective medical evidence support Dr. Zwiefach’s evaluations?

The ALJ found that the record contained no such corroborating evidence. Benton points to the narratives included with Dr. Zwiefach’s evaluations and with treatment notes from her therapist. She contends that this shows that the ALJ’s finding is not supported by substantial evidence. The Commissioner argues that Dr. Zwiefach’s filling out the Mental RFC Assessment in January 1999 was not “objective medical evidence” because he had not personally seen Benton for more than a year. His opinion, the Commissioner argues, must have been based upon his stale recollections of Benton, as well as Benton’s self-reported symptoms and limitations. And, because Benton was found not to be credible, these statements would likewise not be credible.

The Commissioner fails to consider that Dr. Zwiefach’s evaluation was largely informed by his continuing consultations with Benton’s therapists. As Benton notes, the record contains ample evidence of such consultations and reports. The question of whether Dr. Zwiefach could provide a valid assessment based on his collecting information from the direct service providers under his supervision is essentially the same as the determination of whether he is or is not a treating source.

3. Were Dr. Zwiefach’s qualifications and specialization evident in the record?

The ALJ found that Dr. Zwiefach’s opinions warranted less weight than usually accorded a treating psychiatrist because his qualifications and areas of specialty are not evident from the record. But the record shows that Benton identifies him as her treating psychiatrist, his records and correspondence come from Kaiser’s Department of Psychiatry and Addiction medicine, and he is documented as prescribing Elavil and Paxil and consulting with her therapists about her treatment. If the ALJ was not sure as to Dr. Zwiefach’s background, he had the opportunity and duty to develop the record.

4. If Dr. Zwiefach was the treating physician, should his opinion have been given controlling weight?

The Commissioner notes that even if Dr. Zwiefach were accorded the status of a treating physician, his opinion would not necessarily have been conclusive. Thomas v. Barnhart, 278 F.3d 947, 956 (9th Cir.2002). Where, as here, the record contains conflicting medical evidence, the ALJ is charged with determining credibility and resolving the conflict. Id. at 956-57 . The ALJ may reject the opinion of a treating physician in favor of the conflicting opinions of an examining physician if the ALJ makes findings setting forth specific legitimate reasons based on substantial evidence in the record. Id. at 957 .

The Commissioner implies that the ALJ’s summary of Dr. Engelhorn’s report accounted for the possibility that Dr. Zwiefach should be accorded treating physician status. This is untrue. The ALJ determined that neither the putatively stale opinions of Dr. Zwiefach, nor those of anyone else from Benton’s treatment team at Kaiser, would be accorded treating physician status. He then credited Dr. Engelhorn’s opinions (despite characterizing them as those of a non-examining, non-treating medical source as per SSR 96 — 6p) as being supported by evidence in the case record as a whole.

5. Did Dr. Zwiefach’s Mental RFC Assessment indicate disability?

The vocational expert testified that a claimant with the limitations Dr. Zwiefach identified in his assessment would not be able to sustain competitive employment. Benton would not lose at step 4, and it would seem that from the vocational expert’s testimony that she would prevail at step 5 as well.

6.Conclusions regarding Dr. Zwie-fach.

In summary, we conclude we should remand for the ALJ to determine whether Dr. Zwiefach’s treatment relationship with Benton was “typical for her condition” as per § 404.1502. If he finds that it is, he should accord it “treating source” status under § 404.1502, either individually or as representative of Benton’s treatment team at Kaiser. He should develop the record if necessary to explore the qualifications and credibility of the Kaiser treatment team. He should then resolve the conflict of evidence in the record between Dr. Zwiefach and Dr. Engelhorn, redo step 4 of the analysis, and conduct step 5 should it become necessary.

D. Should Benton’s subjective complaints have been credited?

Although the ALJ found Benton had “severe physical and psychiatric impairments,” he nonetheless found Benton’s claims of disabling pain and limitations not credible. Under the test of Cotton v. Bowen, 799 F.2d 1403, 1407 (9th Cir.1986), Benton had to produce objective medical evidence of her impairments and show that the impairments could reasonably be expected to produce some degree of the alleged symptoms. The ALJ could then reject her testimony only upon (1) finding evidence of malingering, or (2) expressing clear and convincing reasons for doing so. The ALJ found no evidence of malinger ing, and so was required to state which symptom testimony he found not credible with enough specificity to allow a reviewing court to confirm that the testimony was rejected on permissible grounds and not arbitrarily. This he has not done. Here, given the obvious serious physical impairments, Benton’s claim of pain and physical limitations should have been credited. See Bunnell v. Sullivan, 947 F.2d 341, 344 (9th Cir.1991) (en banc); Social Security Ruling 96-7p.

CONCLUSION

The crux of this case is an issue of first impression at the appellate level in this circuit: Do emerging patterns of medical treatment that make a psychiatrist responsible for prescribing and monitoring medication, but leave most of the direct patient contact to others within a treatment team, allow the psychiatrist to be a “treating source” either on his own behalf or on that of the treatment team? We conclude that they do.

As to any particular psychiatrist, the question can be comfortably addressed within the confines of § 404.1502. In reaching a finding regarding whether Dr. Zwiefach should have been accorded “treating source” status, the ALJ must explore whether his treatment relationship, individually and as a representative of a treatment team, was consistent with accepted medical practice for the type of treatment required for Benton’s medical condition. If the ALJ finds that Dr. Zwiefach warrants such status and that his opinion is not out-weighed by that of Dr. Engelhorn, he should authorize payment of Benton’s disability benefits. The district court is reversed. It is directed to vacate the Commissioner’s ruling and remand to the ALJ with instructions.

REVERSED AND REMANDED WITH INSTRUCTIONS.

. For this reason, the district court erred in stating that the ALJ's reasoning "that the record does not support giving Dr. Zwiefach’s opinion any greater weight than the other consulted physicians ... is supported by the evidence because there is no indication that he possessed more expertise than the other psychiatrist who had presented a medical assessment.” If Dr. Zwiefach is a treating physician, the primacy of his opinion derives from his superior vantage compared to a non-treating physician, even apart from any superior credentials. . A review of cases citing Matney shows that this portion of the opinion has come to stand for the proposition that “an ALJ need not accept a treating physician’s opinion if it is conclusory and brief and unsupported by clinical findings.” Tonapetyan v. Halter, 242 F.3d 1144, 1149 (9th Cir.2001). Interestingly, it is assumed that Dr. Cookson was a treating physician. . Benton argues that where the Commissioner fails to provide adequate reasons for rejecting the opinion of a treating doctor, that opinion is credited as a matter of law. Hammock v. Bowen, 879 F.2d 498, 502 (9th Cir. 1989). Contrary to Benton’s urging, this does not resolve the case. As noted above, the opinion of even a treating physician can be overcome by substantial contradictory evidence.