Shelley McGoffin appeals from an order entered by a United States Magistrate pursuant to 28 U.S.C. § 636(c)(1) and (3) affirming the decision of the Commissioner of Social Security to deny her application for disability benefits under the Social Security Act. Ms. McGoffin contends on appeal that the Administrative Law Judge (ALJ) did not properly evaluate her treating physician's opinion, did not properly apply the regulations dealing with an applicant's drug and alcohol dependency, did not support his opinion with substantial evidence, and did not properly evaluate her credibility. We reverse and remand for further proceedings.

* Ms. McGoffin filed her application for disability insurance benefits on May 29, 1996. At that time she was twenty years old and had a history of mental illness. On March 31, 1995, she was admitted to Tulsa Regional Medical Center, where she was diagnosed with a major depressive disorder and alcohol dependence. App. vol. II, at 83. She was discharged on April 3, 1995, and readmitted April 10 on an emergency basis when she took an overdose of motion sickness pills. Id. at 97.

On March 25, 1996, she was admitted to Monarch, Inc., a residential chemical dependency treatment program. She left Monarch on May 14, 1996, and was admitted directly to Eastern State Hospital because she appeared suicidal. Id. at 104. She was diagnosed at Eastern State with dysthymic disorder,1 polysubstance abuse, and borderline personality disorder.2 She had mutilated herself with cuts on her arms and abdomen, stating that the pain felt good, and that she liked it because it was the only time she felt anything. She also reported "thoughts of harming people when they make her angry and [that she] threw knives at a co-worker about two months ago." Id. at 106. Eastern State gave her a guarded prognosis, and prescribed Valproic Acid and Prozac. Her psychiatric evaluation while at Eastern State described her liabilities as "a long history of alcohol and drug abuse, poor impulse control, has a tendency to curse, scream, go into a rage, throw knives and as a result she is unable to hold a job." Id. at 110.

She was discharged July 7, 1996, to a controlled residential treatment program with Zarrow/Parkside. She lived at the Zarrow Center from July 1, 1996, to August 14, 1996. She was again diagnosed with dysthymic disorder/psychotic disorder, polydependence, and borderline personality disorder. In her initial assessment, the staff noted that she had strong borderline personality disorder traits. Id. at 220. While at Zarrow she was prescribed Mellaril, Trazodone, Paxil, Valproic Acid, Prozac, Thioridazine, and Carbamazepine. In August she was discharged from Zarrow into a supported, transitional living program. Her referral into this program stated that she needed a supported housing situation with some degree of supervision, and that her previous attempts to live independently had been unsuccessful. While in this program she was in treatment at Parkside four days a week, six hours a day.

On February 12, 1997, Ms. McGoffin attempted suicide by overdosing on Mellaril and Paxil. She was admitted to Parkside on February 19, after the court committed her for twenty-eight days. She was subsequently discharged to Zarrow Center and attended the Parkside partial hospitalization program. On February 24, she was diagnosed by Parkside with major depressive disorder, recurrent, moderate, dysthymic disorder, alcohol, cannabis and amphetamine abuse, and borderline personality disorder. She was prescribed Trazodone, Valproic Acid and Paxil.

Ms. McGoffin returned to full-time employment in August, 1999, although she continued in therapy at Parkside. She claims disability benefits for a closed period, from her alleged onset date of March 23, 1996, to August 31, 1999.

Ms. McGoffin's request for benefits was denied initially and on reconsideration. She then sought a hearing before an ALJ, which was held on September 9, 1997, during the period she was in treatment at Parkside in the transitional living program. Ms. McGoffin was represented by counsel and testified on her own behalf. Also testifying were David Sumter, her case manager in the Parkside program, and a vocational expert.

The ALJ held that Ms. McGoffin was disabled at step three of the sequential evaluation process for mental impairments set out in 20 C.F.R. § 404.1520a. In reaching this determination, the ALJ considered all of Ms. McGoffin's impairments including her substance addiction. The ALJ then applied the Contract with America Advancement Act of 1996, Public L. No. 104-121, 110 Stat. 847 (enacted March 29, 1996). Under this provision, "[a]n individual shall not be considered to be disabled for purposes of this subchapter if alcoholism or drug addiction would (but for this subparagraph) be a contributing factor material to the Commissioner's determination that the individual is disabled." 42 U.S.C. § 423(d)(2)(C). The ALJ found that Ms. McGoffin was able to voluntarily control her substance abuse, and that it was a contributing factor material to her disability determination. The ALJ found that "[t]he record demonstrates an intense exacerbation of the claimant's symptoms when she is under the influence of her substance addiction disorder," App. vol. II, at 25, and that absent the effects of the substance abuse disorder her other mental disorders were not severe impairments at step two.

In reaching this conclusion, the ALJ gave no weight to the assessment of Ms. McGoffin's work-related abilities made by her treatment team when she was in the Parkside transitional living program. This assessment was prepared shortly before the hearing by her case manager, David Sumter, and signed by her treating psychiatrist, Richard Luc, M.D. The assessment concluded that Ms. McGoffin was markedly or severely limited in, for example, her ability to understand and remember detailed instructions; her ability to sustain concentration, perform activities within a schedule, sustain an ordinary routine without supervision, or complete a normal workday without interruption from psychologically based symptoms; and her ability to respond appropriately to criticism from supervisors or to changes in a work setting. Id. at 253-55. The assessment then added the following comment:

It should be noted that the preceding ratings are based on observation of the client in a structured, controlled treatment environment; obviously an environment that does not compare with the general work environment. It is the opinion of the treatment team that this client would function at a much lower level in a work environment. It is also the team's opinion that any past substance abuse engaged in by the client did not necessarily contribute to her mental illness, but was, rather an attempt to self-medicate symptoms of the mental illness.

Id. at 256 (emphasis added). Although the ALJ acknowledged Mr. Sumter's testimony that Dr. Luc agreed with the assessment, the ALJ expressed doubt as to whether the assessment expressed Dr. Luc's "own, singular opinion." Id. at 23. The ALJ also faulted the assessment for distinguishing between Ms. McGoffin's mental illness and her substance abuse because such a distinction was not often made in Ms. McGoffin's many diagnoses.

Finally, the ALJ stated his belief that the assessment was not valid because the treatment team, and Mr. Sumter in particular, had been involved in assisting Ms. McGoffin and that such activities signified a certain advocacy posture which colored the assessment. The ALJ further concluded that to the extent the testimony of Ms. McGoffin and Mr. Sumter tended to show that absent her substance abuse her mental impairments were severe, "such testimony, in light of all of the evidence, including the medical exhibits, is deemed not sufficiently credible to support a finding of disability under current criteria." Id. at 26.

On appeal, Ms. McGoffin contends that the ALJ erred in rejecting the opinion of her treating physician, Dr. Luc, and that the ALJ's various findings were not supported by substantial evidence. We turn first to the ALJ's treatment of the opinion of Dr. Luc.

"An ALJ is required to give controlling weight to a treating physician's well-supported opinion, so long as it is not inconsistent with other substantial evidence in the record." Drapeau v. Massanari, 255 F.3d 1211, 1213 (10th Cir.2001) (citing 20 C.F.R. § 416.927(d)(2)); see also 20 C.F.R. § 404.1527(d)(2). In this case, the ALJ refused to give any weight to the assessment by Ms. McGoffin's treating physician, Dr. Luc, for three reasons. First, despite the fact that Dr. Luc's signature appears directly below the assessment, the ALJ expressed doubt that the assessment was actually that of Dr. Luc. In so doing the ALJ rejected the undisputed testimony of another member of the treatment team, Mr. Sumter, that Dr. Luc reviewed and agreed with the assessment.

"In choosing to reject the treating physician's assessment, an ALJ may not make speculative inferences from medical reports and may reject a treating physician's opinion outright only on the basis of contradictory medical evidence and not due to his or her own credibility judgments, speculation or lay opinion." Morales v. Apfel, 225 F.3d 310, 317 (3d Cir.2000) (internal quotations omitted) (emphasis added). Although we may not second-guess an ALJ's credibility judgments, such judgment by themselves "do not carry the day and override the medical opinion of a treating physician that is supported by the record." Id. at 318. In this case, the ALJ's unfounded doubt that Dr. Luc agreed with the assessment he signed, in the face of unrefuted evidence to the contrary, was error. At the least, if the ALJ believed that the matter was open to question, he had an obligation under the applicable regulations to obtain additional information from Dr. Luc before rejecting the report outright. The regulations state in this regard:

We will seek additional evidence or clarification from your medical source when the report from your medical source contains a conflict or ambiguity that must be resolved, the report does not contain all the necessary information, or does not appear to be based on medically acceptable clinical and laboratory diagnostic techniques.

20 C.F.R. § 404.1512(e)(1) (2001).

The ALJ also faulted Dr. Luc's assessment because it distinguished between Ms. McGoffin's mental illness and her substance abuse although the prior medical reports rarely did so.3 Fairly read, the many medical diagnoses of Ms. McGoffin's condition simply list her mental disorders and her substance addiction and do not distinguish between them in terms of their effect, no doubt because most of these diagnoses were made before the new law in 1996 made such a distinction critical. This fact alone, however, does not preclude Dr. Luc from making such a distinction as her treating physician. We also point out that shortly after the law was amended to provide that a claimant could not be disabled by substance abuse alone, the Social Security Administration sent out a teletype on applying the new law. This teletype stated that "[w]hen it is not possible to separate the mental restrictions and limitations imposed by [drug and alcohol abuse] and the various other mental disorders shown by the evidence, a finding of `not material' would be appropriate." App. vol. I, at 44. In other words, the agency directed that if the effects of a claimant's mental illness could not be separated from the effects of substance abuse, the abuse would be found not to be a contributing factor material to the disability determination.

Finally, the ALJ rejected Dr. Luc's assessment because he felt it signified "a certain advocacy posture." App. vol. II, at 24. This holding, too, was error. We held years ago that an ALJ's assertion that a family doctor naturally advocates his patient's cause is not a good reason to reject his opinion as a treating physician. See Frey v. Bowen, 816 F.2d 508, 515 (10th Cir.1987). Here, as in Frey, the ALJ's rejection of Dr. Luc's assessment on the basis of advocacy is a mere "conclusory statement" that contradicts our rule on the weight to be given the report of a treating physician, "without suggesting some exceptional basis in the facts of this case." Id. Unlike the ALJ, we do not find it exceptional that the treatment team for a patient in a transitional living program assists the patient in making her social security disability claim.

In view of our conclusion that the ALJ erred in rejecting the assessment of Dr. Luc, we must remand this case for further consideration. Before we do so, however, we address Ms. McGoffin's argument that several of the ALJ's findings material to his disability determination are not supported by sufficient evidence.

The ALJ made the critical finding that if Ms. McGoffin were to cease her substance abuse, her remaining mental illness would not be severe enough to preclude work activity. In support of this finding, the ALJ expressly relied on the assessment of a consulting physician, Dr. George Blake, who examined Ms. McGoffin only once, on July 18, 1996, over a year prior to the hearing and while she was living at the Zarrow Center in a controlled residential treatment program. Dr. Blake reviewed her records from Eastern State and from Zarrow. The ALJ specifically relied on Dr. Blake's statements that she was currently clean from substance abuse and that "[p]retty much her cognitive abilities were intact based on this mental status. She related fairly well. The medication seems to have improved her control of her rapid mood swings." App. vol. II, at 124.

This court has long held that "findings of a nontreating physician based upon limited contact and examination are of suspect reliability." Frey, 816 F.2d at 515. Moreover, we do not believe that Dr. Blake's report will bear the weight placed upon it by the ALJ. While it likely was true that Ms. McGoffin was clean and sober while she was in the residential treatment program at Zarrow Center, it is also true that she was in a structured environment there, highly medicated, and in intensive therapy, factors that also presumably contributed to her alleged improved condition. Significantly, Dr. Blake did not express an opinion on the nature of Ms. McGoffin's cognitive abilities were she to be in an independent work environment. Moreover, the medical records from Ms. McGoffin's participation in the treatment program made by her treating physician state that as of July 29, 1996, shortly after her examination by Dr. Blake, she was still having visual hallucinations, hearing voices, and suffering decreased concentration, and that her "psychotic and affective sx's continue[;] Personality disorder continues." Id. at 170. Accordingly, when the record is viewed as a whole, we are not persuaded Dr. Blake's assessment constitutes substantial evidence that Ms. McGoffin's mental disorders were not disabling absent her substance dependence.

Finally, Ms. McGoffin contends the ALJ erred in concluding that she was not a credible witness without identifying which testimony he did not find credible or the grounds upon which he had reached that determination. "Credibility determinations are peculiarly the province of the finder of fact, and we will not upset such determinations when supported by substantial evidence." Kepler v. Chater, 68 F.3d 387, 391 (10th Cir.1995) (internal citations and quotations omitted). Nonetheless, "[f]indings as to credibility should be closely and affirmatively linked to substantial evidence and not just a conclusion in the guise of findings." Huston v. Bowen, 838 F.2d 1125, 1133 (10th Cir.1988) (footnote omitted); see also Kepler, 68 F.3d at 391. The ALJ stated he found Ms. McGoffin not credible to the extent that her testimony conflicted with his conclusion that her mental illness alone was not disabling. The ALJ did not, however, explain and support with substantial evidence which of her testimony he did not believe and why.

We REMAND this case to the district court with instructions to remand the case to the Commissioner for further proceedings in light of this opinion.