Appellant, Ruth Paller, appeals an August 2, 1991, decision of the Board of Veterans’ Appeals (BVA or Board) which found that evidence recently submitted in her claim to entitlement for service connection for the cause of her husband’s death was not new and material. On March 12, 1992, appellant, through counsel, filed a brief. On June 16, 1992, the Secretary of Veterans Affairs (Secretary) filed a motion for summary affirmance, for acceptance of motion in lieu of a brief, and for a stay of proceedings. Appellant filed a motion on June 26, 1992, for an extension of time to respond to the Secretary’s motion for summary affirmance. On August 17, 1992, *537appellant, through counsel, filed a response to the Secretary’s motion for summary af-firmance. This Court vacates and remands the decision to the BVA for adjudication consistent with this opinion.
PACTS
This action arises from the death of World War II veteran, Edward J. Paller, on September 23, 1985. The veteran had active service from June 1943 to January 1945. The veteran was given a 50% rating for service-connected psychoneurosis, effective January 1945. From 1947 to 1977, the schedular rating was reduced to 30%. The rating was then increased to 50%, and then to 100% in November 1978, where it remained until the veteran’s death. The veteran died from cardiopulmonary arrest due to myocardial infarction. At the time of the veteran’s death, the sole service-connected "disability was mixed-type psychoneurosis with anxiety and depression.
Appellant, widow of the veteran, made a claim for dependency and indemnity compensation, alleging the service-connected disability as a substantial cause of his death. See 38 C.F.R. § 3.5 (1991). The BVA denied the claim in November 1986. The BVA found that Mr. Paller’s psychiatric disorder did not cause or contribute substantially or materially to his death, and that his death was caused by a cardiopulmonary arrest due to myocardial infarction.
Appellant thereafter submitted various additional materials in an effort to show that new and material evidence existed to support the reopening of her claim for service connection regarding her husband’s death. On May 1, 1989, the BVA issued a decision stating that the evidence submitted since the 1986 decision of the BVA was essentially duplicative of previously considered evidence. The additional evidence submitted included duplicates of correspondence from the veteran’s private doctors, duplicates of hospital reports, copies of prescriptions for medication, various articles pertaining to stress and stress-related disorders, and the transcript of an Oprah Winfrey segment pertaining to treatment in mental hospitals. Also furnished was a December 1987 letter from Dr. Durkin, a private psychiatrist who treated the veteran. Dr. Durkin concluded that the veteran’s chronic anxiety state produced an irregular heart rate (arrhythmia), which resulted in the veteran’s death. Dr. Durkin furnished a copy of a medical paper entitled, “Ventricular Tachyarrhythmia Associated with Psychological Stress.” The BVA stated that this evidence was insufficient to establish that the veteran’s death was due to a disability related to his period of active service, or causally related to the service-connected psychiatric disorder.
Appellant again submitted additional evidence in an attempt to reopen her claim. The evidence included the article entitled “Ventricular Tachyarrhythmia Associated with Psychological Stress,” which was previously of record, and duplicates of treatment records previously submitted. The additional evidence also included a statement from Dr. Durkin, dated June 1991, and a March 7, 1991, statement from Dr. Byron, a private psychiatrist who treated the veteran. In an August 2, 1991, decision, the BVA stated that the letter from Dr. Durkin was obviously new, but was essentially unchanged from statements he had previously made that were of record prior to the May 1989 decision. Therefore, Dr. Durkin’s June 1991 letter was not “new and material.”
Dr. Byron’s letter stated that the veteran’s “longstanding anxious [sic] disorder producing symptoms over a period of forty years was certainly a stressor that affected cardiovascular status,” and that, with reasonable medical certainty, it was Dr. Byron’s opinion that the veteran’s “chronic emotional stress and anxiety requiring intensive treatment, including electroconvul-sive therapy, directly contributed to his death on September 23, 1985, of cardiopulmonary arrest due to myocardial infarction.” The BVA found the letter from Dr. Byron to be new, but discounted the letter as not being material since the letter was a rearticulation of an opinion (Dr. Durkin’s) that the Board had earlier rejected. The BVA went on to state that both doctors had departed from their role as healers and *538clinicians and assumed the task of advocates. The BVA found the two letters to be highly conclusory, and as such to lack cogency. In the BVA’s judgment, appellant’s argument was speculative when previously made and remained so in the August 2, 1991, appeal. It is the August 2, 1991, BVA decision that is the subject of this appeal.
ANALYSIS
The BVA’s conclusion in this case that the evidence submitted since the Board’s May 1989 decision denying entitlement to service connection for the cause of the veteran’s death is neither new nor material is a conclusion of law, which this Court reviews de novo. See Smith v. Derwinski, 1 Vet.App. 178, 180 (1991), Colvin v. Derwinski, 1 Vet.App. 171, 174 (1991). Appellant attempted to reopen her previously adjudicated case on the basis of “new and material evidence.” See 38 U.S.C. § 5108 (formerly § 3008); 38 U.S.C. § 7104(b) (formerly § 4004(b)). “New evidence” is evidence which is not “merely cumulative” of other evidence in the record. Colvin, 1 Vet.App. at 174. “Material evidence” is “relative and probative of the issue at hand.” Id. However, not every piece of new evidence, even if relevant and probative, will justify a reopening. Id. In order to justify the reopening of a case based on new and material evidence, “there must be a reasonable possibility that the new evidence, when viewed in context of all the evidence, both old and new, would change the outcome.” Id.
This Court has set forth a two-part test for determining whether evidence is “new and material” for purposes of reopening a claim. See Manio v. Derwinski, 1 Vet.App. 140, 145 (1991). The BVA must first determine whether the evidence submitted is “new and material.” Id. Second, if the evidence is new and material, then the case must be reopened and the BVA must evaluate the appellant’s claim in light of all the evidence, both new and old. Id.
The BVA was undoubtedly correct in finding Dr. Durkin’s statement cumulative of the statement he had previously made. It was, therefore, not “new” evidence within the meaning of the statute. The BVA also found Dr. Byron’s statement to be merely a rearticulation of Dr. Dur-kin’s opinion and hence not material. The Court finds this to be error. While it is true that Dr. Durkin’s opinion had been previously considered, and rejected, we find the second opinion by Dr. Byron corroborating Dr. Durkin to be relevant and probative, with a reasonable possibility of changing the outcome, and hence, material. It is the corroborative nature of the opinion that renders it so. Moreover, in a situation such as is presented here, where there is an apparent split in scientific opinion, corroboration is particularly important and cannot be rejected as merely cumulative. This does not mean that a claimant can endlessly reopen a case “a doctor at a time.” There will be a point reached, and probably fairly quickly, where it can be Said that, all things being equal, the evidence being proffered has been fairly considered and that further rearticulation of already corroborated evidence is, indeed, cumulative. This Court will, of course, necessarily deal with that issue on a case-by-case basis.
The Board also committed error in rejecting out of hand the medical opinion proffered because it was “conclusory and as such lacked cogency. The question of the effect that psychiatric illness has upon the development of organic cardiovascular disease is at best unknown.” No medical evidence to support this, also conclusory, statement is cited. This Court understands that there is medical expertise at the Board. We respect that fact but iterate, as we did in Colvin, 1 Vet.App. at 175, that when the Board reaches a medical conclusion it must cite to medical evidence of record to support that conclusion. Otherwise, there is no way for the veteran to know what the medical basis of the decision was, nor for this Court to review properly the evidentiary basis of the decision. Thus, when the medical knowledge of the Board is such that there is a reason to be skeptical of the medical evidence being offered by the claimant, the Board is always free, and should, seek “an advisory opinion *539... or [quote] recognized medical treatises in its decision that support its ultimate conclusions.” Id. at 175; Hatlestad v. Derwinski, 3 Vet.App. 213, 217 (1992).
The decision of the BVA is VACATED and REMANDED for adjudication consistent with this opinion.