MEMORANDUM

The Commissioner of Social Security (the “Commissioner”) appeals the district court’s application of the credit-as-true rule to reverse an administrative law judge’s (“ALJ’s”) order denying Disability Insurance Benefits and Supplemental Security Income to Plaintiff-Appellee Julie Trnavsky (“Trnavsky”) and to remand for the immediate payment of benefits. We affirm.

We review de novo the district court’s decision to reverse the ALJ’s disability determination, see Garrison v. Colvin, 759 F.3d 995, 1010 (9th Cir.2014), but we review the corollary decision to remand for immediate payment of benefits only for abuse of discretion, see Harman v. Apfel, 211 F.3d 1172, 1174 (9th Cir.2000). Under the “Harman test,” a district court may remand for immediate award of benefits only “if (1) the ALJ has failed to provide legally sufficient reasons for rejecting [a treating physician’s opinion], (2) there are no outstanding issues that must be resolved before a determination of disability can be made, and (3) it is clear from the record that the ALJ would be required to find the claimant disabled were [the treating physician’s opinion] credited.” Benecke v. Barnhart, 379 F.3d 587, 593 (9th Cir.2004) (citing Harman, 211 F.3d at 1178 ). If the first element is met, the district court may credit the rejected evidence as true as a matter of law (the “credit-as-true” doctrine) and analyze the next two elements of the Harman test under that assumption. Benecke, 379 F.3d at 594 .

1. Harman Step One. The Commissioner expressly waived any challenge to the district court’s determination (on Harman Step One) that the ALJ failed to provide adequate reasons for rejecting Dr. Holland’s opinion. We therefore find that issue waived. See Burrell v. Colvin, 775 F.3d 1133, 1141-42 (9th Cir.2014). The Commissioner’s argument that the district court abused its discretion in crediting Dr. Holland’s opinion as true therefore lacks merit: Because Harman’s first element was met, the district court acted within its discretion in crediting Dr. Holland’s opinion as true. See Benecke, 379 F.3d at 593 . Moreover, an adverse credibility finding as to the claimant does not justify rejecting a treating physician’s opinion unless (a) that opinion was based “to a large extent” on the claimant’s unreliable self-reports, and (b) the physician did not independently evaluate the claimant’s symptoms and make medical findings. Compare Tommasetti v. Astrue, 533 F.3d 1035, 1037, 1040-41 (9th Cir.2008), with Benecke, 379 F.3d at 594-96 . Here, Dr. Holland’s medical opinion was expressly based on her independent medical assessment, on the corroborating diagnoses of other specialists, and on Trnavsky’s well-documented mental health history.

Nor does the fact that Dr. Holland provided primarily physical treatment provide a legally cognizable basis for rejecting Dr. Holland’s opinion — much less for finding that the district court abused its discretion in crediting Dr. Holland’s opinion as true. Specialization is but one of six factors to be considered in weighing competing medical opinions, and Dr. Holland, as an M.D., was qualified to opine on Trnavsky’s mental health. See Sprague v. Bowen, 812 F.2d 1226, 1232 (9th Cir.1987). Indeed, we have recognized that a primary care physician’s opinion is entitled to “special” weight. See, e.g., Lester v. Chater, 81 F.3d 821, 832-35 (9th Cir.1995).

2. Harman Step 2. The district court likewise did not abuse its discretion in finding, at Harman Step 2, that no outstanding issues remained after crediting Dr. Holland’s opinion as true, even though Dr. Gordy (a non-examining physician) disagreed, with Dr. Holland’s assessment. On closer examination, Dr. Gordy’s opinion does not materially conflict with that of Dr. Holland, Trnavsky’s treating physician. While. Dr. Holland definitively opines that Trnavsky is unable to maintain employment, Dr. Gordy merely offers her “conjecture” that Trnavsky could respond appropriately to work-related people and pressures. The credit-as-true rule permits a reviewing court to “credit as true” an improperly rejected treating physician’s assessment over a non-examining,. non-treating physician’s contrary opinion on the same question. See, e.g., Widmark v. Barnhart, 454 F.3d 1063, 1069 (9th Cir.2006). Particularly in light of the “special weight” afforded to a primary care physician’s opinion, see, e.g., Lester, 81 F.3d at 832-35 , it was not an abuse of discretion for the district court to conclude that Dr. Gordy’s speculative statement did not raise an outstanding issue.

Nor does our recent decision in Dominguez v. Colvin, 808 F.3d 403 (9th Cir.2015) compel a finding that the district court abused its discretion in remanding Trnavsky’s case for immediate award of benefits. The record in Dominguez contained significantly more factual ambiguity than does the record here. Not only was the treating physician’s opinion contradicted by three other physicians, it was contradicted by the treating physician’s own notes. Id. at 408. Moreover, Dominguez had stated a different disability onset date in her second application for disability than she had stated in her first application, and had offered no evidence to explain the discrepancy. Id. at 409 . Given that the determination as to whether “further administrative proceedings would serve no useful purpose” is highly fact-specific, id. at 408 , it does not follow fi’om Dominguez that the district court’s contrary determination in Trnavsky’s case (on a different set of facts) was an abuse of discretion. Indeed, the Commissioner’s argument ignores a key distinction between the posture of Dominguez and the present case, which, given the highly deferential standard of review on appeal, supports a different outcome here. In Dominguez , the district court had remanded the claimant’s ease for further fact-finding', thus, we simply held that this was not an abuse of discretion. Id. at 404, 409 . Conversely here, the district court remanded Trnavsky’s case for immediate payment of benefits, and the Commissioner asks us to find an abuse of discretion. This is a much higher bar, and, for all the reasons set forth herein, we do not find it met here.

3. Harman Step 3. Finally, the district court did not abuse its discretion in holding that if the improperly discredited evidence were credited as true, the ALJ would be required to find Trnavsky disabled on remand. Garrison, 759 F.3d at 1021 . Dr. Holland’s diagnoses of PTSD and depression, combined with Dr. Aviot-ti’s assessment and the uncontroverted evidence of Trnavsky’s job history, demonstrated that Trnavsky’s PTSD and other health issues caused emotional outbursts in the workplace and excessive absenteeism. In turn, a vocational expert called by the government testified that an individual who “decompensates” in the workplace because of PTSD {i.e., the individual’s mental illness manifests because of triggers in the workplace), or who routinely requires more than two days off from work in a month, would be unable to maintain competitive employment. Because this evidence established that Trnavsky’s impairments would rule out competitive employment, the district court did not abuse its discretion in finding that the evidence compelled a finding of disability. See, e.g., Lester, 81 F.3d at 834-35 .

For all of these reasons, we AFFIRM the district court’s reversal and remand for immediate payment of benefits.

This disposition is not appropriate for publication and is not precedent except as provided by 9th Cir. R. 36-3. , This is an issue that we would generally review de novo because it goes to the question whether the remand was proper in the first instance. See Harman, 211 F.3d at 1174 (decisions to affirm or reverse the AU’s determinations are reviewed de. novo, as is the district court's determination to remand the case). . Nor was Dr. Holland’s three-page medical opinion overly brief and insufficiently supported by clinical findings; thus, Bayliss v. Barnhart, 427 F.3d 1211 (9th Cir.2005), does not help the Commissioner. .And contrary to the Commissioner’s assertion, the district court did not affirm the ALJ’s decision to afford great weight to Dr. Gordy’s opinion. Rather, the district court held that, because Dr. Gordy’s testimony that Trnavsky would not get along with a “hard-nosed, non-compassionate supervisor” was purely speculative, the ALJ did not err by failing to include limitations relating to a supervisor's personality in Trnavsky’s Residual Functional Capacity (“RFC”). Thus, there is no internal inconsistency that would require reversal of the district court’s opinion,