Claimant Gary Rose filed an application for Social Security disability benefits on August 13, 1990, alleging chronic fatigue syndrome (CFS), back pain, and a mental condition. After a hearing, an Administrative Law Judge (ALJ) conceded that claimant had a severe impairment or impairments that precluded his return to his former job as a grocery clerk required to do medium-to-heavy work. The ALJ found, however, that despite claimant’s exertional impairments he retained the residual functional capacity to perform sedentary work. The ALJ further found that claimant’s non-exertional impairments (his pain, his mental condition, and the subjective symptoms associated with CFS) did not significantly restrict his capacity to perform the full range of jobs requiring sedentary work. And, moreover, the ALJ received testimony from a vocational expert that, notwithstanding claimant’s impairments, there existed a significant number of sedentary jobs in the economy that claimant could perform, such as marker, sorter, packager, boxer, and carder.

Accordingly, the ALJ applied Rule 201.27 of the Medical-Vocational Guidelines, 20 C.F.R. Part 404, Subpart P, Appendix 2 (the Grid) to reach a finding of not disabled. The ALJ made an alternative finding that claimant was not disabled at step 5 of the sequential evaluation process, 20 C.F.R. § 404.-1520(f), on the ground that the Secretary had demonstrated the existence of jobs in the economy that claimant could perform.

After both the Appeals Council and the district court refused to set aside the ALJ’s decision, claimant appealed to this court. We vacate the Secretary’s final decision and remand for further proceedings.

The Medical Evidence

Claimant alleged back pain resulting from an injury he suffered at work in December 1987. According to claimant, he began to[*15] experience the symptoms associated with CFS in June 1988.

The relevant medical evidence, listed in rough chronological order, can be summarized as follows. On September 29,1988, Dr. Hillier, a treating orthopedist, diagnosed a chronic lumbar radiculopathy, but concluded that claimant “can return to work involving no repetitive heavy lifting.” On November 16,1988, Dr. Hillier stated that claimant “has made a good recovery” from his back problems, and concluded: “He is going to return to work; other than extremely heavy lifting which is not required of his job, I see no reason why he could not return to work.” On March 6, 1989, Dr. Hillier stated that, orthopedically, there was “no sign of any malingering,” and that claimant, from an orthopedic standpoint, “could return to light duty status work at the supermarket.” The doctor noted, nevertheless, that “there seems to be a problem.” Dr. Hillier made no mention of CFS or any of the symptoms associated with it; instead, he confined his findings to claimant’s lumbosacral problems.

In the fall of 1989, the circumstances changed. On October 11, 1989, Dr. Hillier reported that claimant “has developed an unrelated problem of weight loss, low-grade fever and swollen glands. He has been seen by numerous medical doctors and no definitive diagnosis has been made.” On December 19, 1989, Dr. Hillier stated that claimant’s “workup [had become] consistent with a chronic fatigue syndrome.” On July 12, 1990, Dr. Hillier wrote that claimant was bedridden for weeks at a time and “is unable to participate in any kind of exercise because of constant fatigue.”

Dr. Perl, a treating physician, diagnosed claimant’s back problems in 1989. His October 26, 1989 report concluded that claimant “remains totally disabled.” In a July 31, 1990 report, Dr. Perl found claimant “partially disabled.” Both reports were limited to assessments of claimant’s lumbosacral condition; neither report mentioned CFS or its symptoms.

On May 24, 1990, Dr. Chowdri, a treating physician specializing in internal medicine and infectious disease, reported that when he first examined claimant on November 6, 1989, claimant “had generalized malaise, which he found to be quite disabling, recurrent sore throat, and weight loss.” At that time, the doctor found claimant “extremely fatigued and ... not able to work.” Dr. Chowdri’s report indicated that, in a series of later visits (through April 27, 1990), claimant continued to complain of fatigue, low-grade fever, and sore throat. Dr. Chowdri diagnosed CFS. In a report dated September 14, 1990, Dr. Chowdri noted that, in three ensuing visits, claimant had continued to complain of fatigue and “generalized malaise.” Nonetheless, a physical examination proved “unremarkable.” Dr. Chowdri stated that he could not “find any physiological reason why this patient cannot return to work.” On an accompanying form, Dr. Chowdri endorsed his opinion that “this patient can return to work.”

Dr. Wagner, a treating physician, stated in a September 18, 1990 report that claimant’s medical tests were unremarkable; he wrote that “[t]o my knowledge, at this time [claimant] has no major medical illnesses” and “is not disabled_” While Dr. Wagner noted that other doctors had diagnosed CFS, he deemed himself “unqualified to recommend disability on the basis of chronic fatigue syndrome” and suggested that claimant “seek counsel of a subspecialist in infectious disease in this regard.”

In a September 27, 1990 report, Dr. Harris, an internist, reported that he had examined claimant on two occasions (in July and September 1990). Although claimant “described a two year history of malaise and fatigue” to Dr. Harris, a “[g]eneral physical examination was unremarkable.” The physician concluded that claimant “may fit the diagnosis of so-called chronic fatigue syndrome though there are clearly no definitive diagnostic tests.”

On January 7,1991, Dr. Weinstein, a treating physician, noted that although numerous diagnostic tests had been negative or normal, for two years claimant “has been debilitated by intermittent episodes of severe sore throat, low-grade fevers, intermittent diarrhea, severe headaches and disabling fatigue to the point [where] he can’t work.” Dr.[*16] Weinstein “suspect[ed] ... underlying chronic fatigue syndrome.” On February 20,1991, Dr. Weinstein reported that claimant remained “very tired, unable to function very well,” and concluded that “[a]t this time, all is consistent with chronic fatigue.”

On April 9, 1991, Dr. Tosches, a treating neurologist, noted that claimant had a long record of complaining about many of the symptoms normally associated with CFS. Dr. Tosches diagnosed “chronic fatigue immunodeficiency syndrome,” saying that the diagnosis was “documented in this patient’s history and medical notes which support the history.”

On May 22, 1991, Dr. Weiss, a treating physician, reported that claimant had symptoms of “fatigue, nausea, [diminished] concentration, frequent sore throats, dysuria, and diffuse aches,” and that these symptoms “wax & wane, but [are] always present, at times more severe.” He thought claimant was “[o]ften too fatigued to carry out routine tasks of life.”

In addition, two non-examining physicians, after reviewing the medical evidence in the record, both checked boxes on residual functional capacity assessment forms indicating that claimant could lift at least 20 pounds (at least 10 pounds frequently), could stand, sit, or walk six hours, and could climb, balance, stoop, kneel, crouch, and crawl at least occasionally. No other functional limitations were noted. One physician explained this evaluation by stating that “[t]here has been no objective abnormality found to explain the fatigue.” The other physician relied on Dr. Chowdri’s comment “that physical condition does not preclude work.” These two residual functional capacity assessments correspond to a finding that claimant could perform light work. See 20 C.F.R. § 404.1567(b).

There is also some psychiatric evidence in the record. On November 27, 1990, Dr. Schembri, a treating psychologist, diagnosed CFS and secondary depression. In a report of a telephone conversation with Dr. Schem-bri, an HHS official observed that Dr. Schembri said that “claimant is legitimate in [his] disease,” and that the doctor “feels strongly that claimant cannot work. He has no energy. He has been suffering for a long time and did not apply for assistance until Dr. Schembri pressed him.”

Dr. Delgado, a consulting psychiatrist, examined claimant on December 5, 1990. Dr. Delgado was not of much assistance as to CFS. He stated: “This is an individual who has a syndrome which I can’t comment on except to say that nothing has apparently turned up in physical or laboratory studies as far as I can determine.”

Dr. Pereira, a consulting psychologist, examined claimant on July 19, 1991. He noted claimant’s complaints of “fatigue-like symptoms for the past three years,” but found that, “[p]sychodiagnostically,” the only “clear indication ... suggestive of any serious psychopathology” was that claimant “may be experiencing a mild adjustment disorder.”

The Secretary’s Policy Concerning CFS

In section DI 24575.005 of the Secretary’s Program Operations Manual System (1993) (POMS), the Secretary established a policy for the evaluation of claims premised on CFS. This policy states in pertinent part:

Chronic Fatigue Syndrome (CFS), previously known as Chronic Epstein-Barr Virus Syndrome, and also currently called Chronic Fatigue and Immune Dysfunction Syndrome, is a systemic disorder consisting of a complex of variable signs and symptoms which may vary in duration and severity. The etiology and pathology of the disorder have not been éstablished. Although there are no generally accepted criteria for the diagnosis of cases of CFS, an operational concept is used by the medical community. There is no specific treatment, and manifestations of the syndrome are treated symptomatically.

CFS is characterized by the presence of persistent unexplained fatigue and by the chronicity of other symptoms. The most prevalent symptoms include episodes of low-grade fever, myalgias, headache, painful lymph nodes, and problems with memory and concentration. These symptoms fluctuate in frequency and severity and may be seen to continue over a period of many months. Physical examination may be within normal limits. Individual cases[*17] must be adjudicated on the basis of the totality of evidence, including the clinical course from the onset of the illness, symptoms, signs, and laboratory findings. Consideration should be given to onset, duration, severity and residual functional capacity following the sequential evaluation process.

POMS § DI 24575.005 (1993).

To be sure, this particular version of the policy was not promulgated until November of 1993, after the ALJ had issued his decision and the Appeals Council had affirmed it. Yet the previous version of the same section of the POMS set forth the same principles governing the evaluation of chronic Epstein-Barr virus syndrome (CEBV). See Reed v. Secretary of HHS, 804 F.Supp. 914, 918 (E.D.Mich.1992). Although medical authorities have stated that “notwithstanding initial inferences, ... EBV [Epstein-Barr virus] is not the exclusive or even principal cause of Chronic Fatigue Syndrome,” id. at 920 (paraphrasing the Harvard Medical School Health Letter on Chronic Fatigue Syndrome (July 1988)), several courts have noted the close similarity and possible relationship between the two maladies and have treated them more or less as peas in a pod. See, e.g., Cohen v. Secretary of HHS, 964 F.2d 524, 529 (6th Cir.1992); Thaete v. Shalala, 826 F.Supp. 1250, 1251 (D.Colo.1993); Reed, 804 F.Supp. at 918. As the Cohen court wrote:

Due to the close association and suspected causal relationship between the Epstein-Barr virus and chronic fatigue syndrome, the two are sometimes referred to synonymously. Although recent studies suggest that the causal relationship between the Epstein-Barr virus and chronic fatigue syndrome may in fact be more attenuated than initially believed, [the two terms] continue to be used somewhat interchangeably.

Cohen, 964 F.2d at 529. Given this historical background, we conclude that, as a practical matter, the POMS policy we have quoted was in effect for CFS cases at the time the ALJ decided this case. ^[1]

The POMS demonstrates the Secretary’s acceptance of certain propositions concerning the nature and medical diagnosis of CFS. These verities — only recently established by the medical community — have been noted by other courts, see, e.g., Sisco v. Department of HHS, 10 F.3d 739, 744 (10th Cir.1993) (“At this point there is no ‘dipstick’ laboratory test for chronic fatigue syndrome;” the medical community instead uses an “operational” diagnostic procedure, so the disease is “not per se excluded from coverage because it cannot be conclusively diagnosed in a laboratory setting”); Reed, 804 F.Supp. at 920-21 (similar), and form the framework for our decision. ^[2]

The Diagnosis of CFS

The ALJ found that claimant has “possible” CFS. We think that this finding grossly understates the nisi prius roll. The administrative record reveals no genuine issue of medical fact: claimant does suffer from CFS. As we explain below, this conclusion is irresistible.

Dr. Tosches and Dr. Weinstein both diagnosed CFS. Although other doctors stated that they were not equipped to speak defini[*18] tively to whether claimant had CFS, no doctor rejected a diagnosis of CFS. And virtually all the doctors who did not disclaim the ability to assess the matter found that claimant had symptoms fully consistent with CFS. Moreover, from mid-1989 forward, the medical references in the record to symptoms of CFS are strikingly consistent.

Nor is the length of time that passed harmful to claimant’s case. Diagnosing CFS is not sport for the short-winded. “[Bjecause chronic fatigue syndrome is diagnosed partially through a process of elimination, an extended medical history of ‘nothing-wrong’ diagnoses is not unusual for a patient who is ultimately found to be suffering from the disease.” Sisco, 10 F.3d at 745. The absence of definitive diagnostic tests, see POMS, supra; see also Sisco, 10 F.3d at 744, makes it plain that the failure of some doctors to state conclusive diagnoses does not constitute substantial evidence to support a finding that claimant did not suffer from the syndrome. See Sisco, 10 F.3d at 745 (findings of physicians who did not rule out CFS, but “merely expressed an inability to discover an adequate physical explanation for [claimant’s] symptoms,” do not constitute substantial evidence to rebut other physicians’ diagnoses of CFS).

We will not paint the lily. It is common ground that an ALJ is not free to substitute his own judgment for uneontroverted medical opinion. See, e.g., Rosado v. Secretary of HHS, 807 F.2d 292, 293-94 (1st Cir.1986) (per curiam). In this case, uniform medical opinion requires a finding that claimant suffers from CFS.

The Functional Significance of Claimant’s Fatigue

Because the medical evidence bound the Secretary to find that claimant does have CFS, the Secretary had no choice but to conclude that the claimant suffers from the symptoms usually associated with CFS, unless there was substantial evidence in the record to support a finding that claimant did not endure a particular symptom or symptoms. Chief among these symptoms, of course, is “persistent unexplained fatigue.” POMS § DI 24575.005 (1993). The record does not contain any meaningful evidence to support a finding that claimant did not suffer from a significant level of fatigue on a regular basis. Leaving to one side Dr. Perl and Dr. Hillier (in his earlier reports) — both of whom confined themselves to discussing claimant’s back condition — all the other examining physicians’ reports, over a period of more than 18 months, consistently noted (and credited) claimant’s complaints of persistent fatigue. ^[3]

The record also contains reports of two non-examining physicians that failed to note any significant functional limitations resulting from fatigue. But, one relied on Dr. Chowdri’s statement, the other on the lack of any “objective abnormality found to explain the fatigue.” The former constitutes too weak a reed for such reliance. See supra n. 3. And the latter is entitled to no weight; as the POMS makes clear, lack of objective proof is what one may expect in cases of CFS.

We have held that the amount of weight that can properly be given the conclusions of non-testifying, non-examining physicians “will vary with the circumstances, including the nature of the illness and the information provided the expert.” Berrios Lopez v. Secretary of HHS, 951 F.2d 427, 431 (1st Cir.1991) (per curiam) (quoting Rodriguez v. Secretary of HHS, 647 F.2d 218, 223 (1st Cir.1981)); accord Gordils v. Secretary of HHS, 921 F.2d 327, 328 (1st Cir.1990) (per curiam). In some cases, written reports submitted by non-testifying, non-examining physicians cannot alone constitute substantial evidence, see Browne v. Richardson, 468 F.2d 1003, 1006 (1st Cir.1972), although this is not an ironclad rule, see Berrios Lopez, 951 F.2d at 431; Gordils, 921 F.2d at 328.

The deciding factor in this case is “the nature of the illness.” Berrios Lopez, 951[*19] F.2d at 431 (quoting Rodriguez, 647 F.2d at 223). The non-examining physicians relied on what they discerned as a lack of objective findings sufficient to prove the existence of significant fatigue. Given the uncontrovert-ed medical evidence that claimant suffered from CFS, however, blind reliance on a lack of objective findings is wholly inconsistent with the Secretary’s policy in such cases as expressed in the POMS and in other pertinent policy statements. See, e.g., POMS § DI 24575.005 (1993) (continuing that although “[pjhysical examination may be within normal limits,” nevertheless, “[ijndividual cases must be adjudicated on the basis of the totality of evidence”).

Furthermore, the medical evidence establishes that claimant possesses a medical condition — CFS—that can reasonably be expected to produce the alleged fatigue. The question here is the extent to which claimant’s fatigue in fact restricts his residual functional capacity. Such an inquiry — into the functional implications of a claimant’s subjective symptoms — “is the kind of inquiry for which on-the-spot examination and observation of claimant might ordinarily be thought important.” Berrios Lopez, 951 F.2d at 432. The subjective severity of a claimant’s fatigue associated with CFS is not something readily evaluated on an algid administrative record.

Under the particular circumstances of this case, we hold that, even if the non-examining physicians’ notations can be read to suggest that claimant’s fatigue did not significantly affect his functional capacity, these notations, without more, could not support the ALJ’s finding to that effect. And because this comprises the only evidence in support, we conclude that the ALJ’s finding is not supported by substantial evidence.

Application of the Grid

The Grid is based on a claimant’s exertional capacity and can only be applied when claimant’s non-exertional limitations do not significantly impair claimant’s ability to perform at a given exertional level. See Sherwin v. Secretary of HHS, 685 F.2d 1, 2-3 (1st Cir.1982), cert. denied, 461 U.S. 958, 103 S.Ct. 2432, 77 L.Ed.2d 1318 (1983). Since the medical evidence in this case compelled a finding that claimant’s fatigue resulting from CFS did significantly impair claimant’s ability to perform even sedentary work, the ALJ erred in relying on the Grid to reach a finding of “not disabled.”

The Vocational Testimony

The ALJ based his determination that claimant was not disabled not only on the Grid but also on the testimony of a vocational expert who, in response to the ALJ’s hypothetical, opined that claimant could perform a number of jobs. The ALJ’s hypothetical, however, impermissibly omitted any mention of a significant functional limitation arising from the fatigue symptoms associated with CFS. Because the ALJ’s hypothetical assumed that fatigue did not pose a significant functional limitation for the claimant, and because the medical evidence did not permit that assumption, the ALJ could not rely on the vocational expert’s response as a basis for finding claimant not disabled. See, e.g., Arocho v. Secretary of HHS, 670 F.2d 374, 375 (1st Cir.1982).

Conclusion

We need go no further. For the reasons we have stated, the judgment of the district court is vacated and the case is remanded to the district court with instructions to remand to the Secretary for further findings and/or proceedings not inconsistent with this opinion.

It is so ordered.