OPINION

Petitioner Sadie M. Hollis challenges the final decision of the Commissioner of Social Security, affirmed by the District Court, that she was not disabled during the period July 16, 1996 through June 30, 1998. Hollis contends that the administrative law judge (ALJ) failed adequately to justify his findings that Hollis’s impairments did not meet or equal the criteria of an impairment listed in 20 C.F.R. § 404, subpt. P, App. 1 (a “Listed Impairment”) and that Hollis was able to perform her past work until July 1, 1998. Because the Commissioner’s decision was supported by substantial evidence, we will affirm.

I.

As we write only for the parties, we recite only the essential facts. Hollis filed for Social Security disability benefits in March 1997, alleging that she had been disabled since July 12, 1996 — also the date on which the data entry facility at which she had been employed closed down. Hollis alleges that her diabetes, diagnosed in 1994, caused problems with her hands and feet since late 1995. She testified that her arms, legs and vision bothered her and restricted her movement, and that she could not lift more than ten pounds.

Several medical reports bear on Hollis’s capabilities prior to July 1998. First, a one page report submitted by Hollis’s treating physician, Dr. Sylvia Pagulayan, in August 1999 states that Hollis was diagnosed with peptic ulcer disease and Type II diabetes mellitus in 1994. The report also describes symptoms of peripheral neuropathy and carpel tunnel syndrome and concludes that Hollis cannot work, but it does not state when her disabling condition began. Dr. Pagulayan’s report was apparently prepared using medical notes from her regular examinations of Hollis since 1984 or 1985. A second report that refers to the period prior to July 1998 is that of Dr. Jean Messihi of New Jersey’s Disability Determination Services. Dr. Messihi performed a number of tests on Hollis on June 19, 1997 and reported no severe impairments. In particular, he found that Hollis possessed “normal function of the hands.” Third, radiologist Dr. Anthony F. Tramontana prepared a report based on a roentgen examination of Hollis’s chest and lumbosacral spine on June 26, 1997. Fourth, Dr. Raquel Eleuteri of New Jersey’s Disability Determination Services submitted a report based on her examination of Hollis on November 10, 1998. Dr. Eleuteri found that Hollis’s hand and wrist were in severe pain and restricted Hollis’s movements. The ALJ relied on Dr. Eleuteri’s report and other evidence to conclude that Hollis was disabled from July 1998 onward; he found there was no medical evidence to support a finding of disability before that time.

*398Hollis argues that the ALJ did not offer any explanation as to how he concluded that her impairments were not the medical equivalent of a Listed Impairment. She further argues that he erred in his assessment of her residual functional capacity prior to July 1, 1998 because he (1) rejected her treating physician’s medical report; (2) did not consult a medical expert about the onset date of her disabilities; (3) rejected Hollis’s subjective reports of pain during this period as incredible; and (4) failed to compare Hollis’s capacities with the demands of her past work.

II.

The Commissioner determines whether a claimant is disabled pursuant to a five-step test. See 20 C.F.R. § 404.1520. Following that five-part test, the ALJ in this case found, first, that Hollis had not engaged in substantial gainful activity, and second, that her impairments were severe within the meaning of 20 C.F.R. § 404.1520(c). At the third step, the ALJ was to determine whether her impairments either met or were equal in severity with a Listed Impairment; he found that they were not. See 20 C.F.R. § 404.1520(d). At the fourth step, he was to determine whether her residual functional capacity was sufficient to perform her past work; he found that until July 1, 1998, it was. See 20 C.F. R. § 404.1520(e, f). As a result, the ALJ did not reach step five of the disability test with respect to the period prior to July 1998. Id.

A. Medical Equivalence

Hollis claims that the ALJ did not properly justify his conclusion that her impairments did not automatically qualify her as disabled at step three. In particular, she argues that the ALJ failed to evaluate the medical evidence as required by 20 C.F.R. § 404.1526(a), failing even to identify which of her impairments he compared to the Listed Impairments. So unsatisfactory was the ALJ’s discussion, contends Hollis, that it is not amenable to review and reversable for that reason under Burnett v. Commissioner of Social Sec. Admin., 220 F.3d 112 (3d Cir. 2000).

In Burnett, this Court vacated a decision whose analysis at step three consisted only of a conclusory statement that the claimant’s impairment did not qualify as a Listed Impairment, without setting forth any reasons that would have allowed for meaningful review. Id. at 120-21. But unlike in Burnett, the ALJ in this case did explain his decision that Hollis’s impairments did not equal a Listed Impairment. Specifically, he found that though she suffered from diabetes and hypertension, in the contested period there was no evidence of any of the following conditions that could render her impairments equivalent to a Listed Impairment: retinopathy or acidosis at least on the average of once every two months, amputation, end organ damage, cerebral vascular pathology, mycardial infarction, blockage in the blood vessels, or substantial adenocarcinoma.

Nor was the ALJ required to consult a medical expert on the question of medical equivalency. Regulations allow but do not require a medical expert to opine on the nature and severity of a claimant’s impairment and whether it equals the requirements of a Listed Impairment. See 20 C.F.R. § 404.1527(f)(2)(iii). The ALJ based his finding as to medical equivalence on medical findings, pursuant to 20 C.F.R. § 404.1526(b). The available medical evidence was adequate and it supported the ALJ’s decision.

B. Capacity To Perform Prior Work

Hollis also claims that the ALJ erred when he found in step four that prior to July 1, 1998, her residual functional capacity was sufficient to perform her past work.

*399First, she argues that he unjustifiably rejected the medical opinion of her treating physician, Dr. Pagulayan. Because treating physicians are “most able to provide a detailed longitudinal picture of a claimant’s medical impairment(s)” and accordingly bring a “unique perspective” to the medical evidence, their opinions are controlling if well-supported and uncontradicted. 20 C.F.R. § 404.1527(d)(2). In this case, had Dr. Pagulayan represented to the Commissioner that Hollis was disabled since her alleged onset date, the ALJ might have concluded that her opinion was contradicted by the report submitted by Dr. Messihi and could potentially have rejected Dr. Pagulayan’s report in favor of Dr. Messihi’s after due consideration of the relevant factors and medical evidence. Id.See also Morales v. Apfel, 225 F.3d 310, 317 (3d Cir. 2000). But in fact Dr. Pagulayan never stated when the disabling symptoms she described began. Even though the ALJ left the record open after the hearing in order to give Hollis an opportunity to obtain medical notes or other evidence from Dr. Pagulayan, Hollis failed to produce any additional documentation of her disability that might have indicated that those symptoms began prior to July 1,1998.

Second, Hollis argues that the ALJ was obliged to consult a medical expert to determine the onset date of her disability. But such consultation is only mandatory when the medical evidence is ambiguous. See SSR 82-30, 1983 WL 31249, at *2-*3; Walton v. Halter, 243 F.3d 703, 709-10 (3d Cir. 2001). There is no evidence on record that suggests Hollis was disabled prior to July 1, 1998. Consultation with a medical expert would nevertheless be required if the absence of evidence suggesting Hollis was disabled was attributable to gaps in her medical evidence.1 Id. But not only are two medical reports from the period prior to July 1998 already on the record, but also, the frequency of Hollis’s visits to Dr. Pagulayan over fifteen years suggests that extensive records are probably in existence, probably in Dr. Pagulayan’s possession. Because it is claimant’s burden at all but the last stage of the five-part test to produce evidence of disability, where there is evidence that complete medical records for a given period are already in existence, an ALJ is not required to engage a medical expert to identify an onset date just because claimant contends that those existing but unproduced records would contradict other medical reports already on the record. See Doak v. Heckler, 790 F.2d 26, 28 (3d Cir. 1986) (holding that plaintiff bears burden in fourth stage of disability test).

Third, Hollis claims that the ALJ inappropriately rejected her subjective testimony regarding her pain. In particular, she claims it is implausible that she became “credible” only after July 1, 1998. Although subjective reports of pain cannot alone sustain a disability claim, the ALJ was indeed obligated to give appropriate consideration to Hollis’s testimony and explain the reason for finding it incredible in part. See 42 U.S.C. § 423(d)(5)(A), 20 C.F.R. § 404.1529(a) (providing that statements about pain or other symptoms will not alone establish disability but agency will consider those symptoms and the extent to which they can reasonably be accepted as consistent with the objective medical evidence). That is what the ALJ did in this case. Hollis’s reading of the ALJ’s decision to imply that Hollis’s credibility as a person magically altered on July 1, 1998 is disingenuous. It is clear that the ALJ found that there was not sufficient evidence of the intensity, frequency *400and duration of Hollis’s alleged symptoms prior to July 1998 to support her subjective complaints, and that the medical evidence suggests her condition would only have been reasonably expected to produce the alleged symptoms to the requisite degree from July 1998 forward. Contrary to Hollis’s suggestion, there is nothing suspicious about the ALJ’s finding that the objective evidence was consistent with her subjective complaints in one period but not another.

Finally, Hollis claims the ALJ failed to compare her residual functional capacities with the demands of her job, as necessary to arrive at a conclusion regarding her ability to perform her past work. See 20 C.F. R. § 404.1520(f). But the ALJ performed precisely that analysis. He observed that Hollis’s position as a data entry clerk did not require exertion in excess of the sedentary level. He acknowledged that it did require bilateral manual dexterity but found that there was no evidence in the record to establish that Hollis was limited in the use of her hands prior to July 1998. It was because she could no longer use her hands for repetitive movements after that time that the ALJ found Hollis was disabled starting July 1, 1998.

Because each of the findings made by the ALJ in arriving at his conclusion that Hollis was disabled since, but not prior to, July 1, 1998 were supported by substantial evidence, we will affirm.