dissenting:
I respectfully dissent because I am unable to agree with the majority’s conclusion *479that the appellant has failed to submit new and material evidence so as to require that the claim be reopened and adjudicated on the basis of all the evidence, new and old, under 38 U.S.C.A. § 5108 (West 1991). See Jones (McArthur) v. Derwinski, 1 Vet.App. 210, 213 (1991); Colvin v. Derwinski, 1 Vet.App. 171, 174 (1991); Manio v. Derwinski, 1 Vet.App. 140, 145 (1991). In so concluding, the Court rejects as not “material” a medical opinion by a treating physician that his patient currently suffers from a psychiatric problem that, based (as the majority concludes) on what the patient recounted to him about in-service treatment and symptoms, is connected to his service 20 years before. For the reasons that follow, I believe that the majority’s analysis is inconsistent with both the record on appeal and this Court’s precedents, and will sow great confusion in the adjudication process throughout the Department of Veterans Affairs (VA).
I. Law on Reopening
Under 38 U.S.C.A. § 7104(b) (West 1991), a claim previously and finally denied by the Board of Veterans’ Appeals (BVA or Board) may not be reopened except in accordance with section 5108. On claims to reopen previously and finally denied claims, the Board must conduct a “two-step analysis” under section 5108. Manio, supra. First, it must determine whether the evidence presented or secured since the pertinent prior final disallowance is “new and material”. Ibid. If it is, the Board must then adjudicate the claim on the basis of all the evidence, both new and old. See Jones, supra. The determination as to whether evidence is “new and material” is a question of law, which this Court reviews de novo under 38 U.S.C.A. § 7261(a)(1) (West 1991). See Masors v. Derwinski, 2 Vet.App. 181, 185 (1992); Jones, 1 Vet.App. at 213; Colvin, supra. The Court recently synthesized the applicable law as follows:
“New” evidence is that which is not merely cumulative of other evidence of record. “Material” evidence is that which is relevant to and probative of the issue at hand and which, as this Court stated in Colvin, supra, ... must be of sufficient weight or significance (assuming its credibility) that there is a reasonable possibility that the new evidence, when viewed in the context of all the evidence, both new and old, would change the outcome.
Cox v. Brown, 5 Vet.App. 95, 98 (1993). See also Justus v. Principi, 3 Vet.App. 510, 513 (1992) (in determining whether evidence is new and material “the credibility of the evidence is to be presumed”).
II. Post-July 1990 Evidence
In the instant case, the majority concludes that no new and material evidence has been submitted since the BVA’s July 1990 prior final disallowance of the veteran’s psychiatric-disorder claim and that, therefore, the Board was not required to proceed to the section 5108 step-two full adjudication of that claim.
I agree with the majority that, of the evidence submitted after the July 1990 BVA denial of the veteran’s claim, item (2) is not “new” and items (1), (3), and (5) are not “material”. I cannot agree, however, with the majority’s conclusion that Dr. Estes’ August 1990 and January 1991 letters are not, taken together, “material”.
A. The Court’s Factfinding
The Court has drawn inferences from the record as to Dr. Estes’ two letters that are not supported by the evidence. As to the Court’s statement that “neither letter contains a current diagnosis of a psychiatric disorder”, ante at 478, Dr. Estes’ August 1990 letter stated: “[The veteran] has had a problem with psychoses for a long period of time and in my opinion it is connected with conditions originating in the service.” R. at 62; Supp.R. at 2. By identifying the veteran’s “problem” as “psychoses”, the letter satisfies the definition of a diagnosis: “1 a: the art or act of identifying a disease from its signs and symptoms[;] b: the decision reached by diagnosis”. WebsteR’s Medical Desk Dictionary 177 (1986). By using the present perfect (“He has had a problem”) and present (“it is connected”) tenses, the letter clearly refers to the veteran’s then-current condition.
*480Second, I also cannot agree with the majority’s conclusion that “there is no indication that [Dr. Estes] formed [his] opinion on a basis separate from appellant’s recitation of his medical and service background”. Ante at 478. In so finding, the majority overlooks the prefatory phrase “in my medical opinion” with which Dr. Estes introduces his August 1990 assertion of a nexus between the veteran’s current disability and his service. The January 1991 statement by Dr. Estes accurately recites several relevant details of the veteran’s in-service January 1964 hospitalization at an Army facility, including both the examiner’s “Diagnosis]” of “Hysteria” and the treatment plan of Thorazine and Nembutal that the examiner prescribed at that time. Compare R. at 10 with R. at 67 and Second Supp.R. at 1. Despite the perfect accuracy with which Dr. Estes’ January 1991 statement recapitulates the January 1964 treatment record, the majority, without acknowledging the congruity of those two records, finds “no indication” that Dr. Estes had reviewed the veteran’s service medical records, and instead finds that Dr. Estes merely parroted the veteran’s “recitation of his medical and service background”. Rather than accept the, at least equally, logical explanation that Dr. Estes had indeed reviewed the veteran’s service medical records (SMRs) and thus formed his opinion based “on independent grounds”, the majority apparently bases its factfinding on the veteran’s ability to recall with flawless precision the details of a single medical treatment episode that had occurred more than two decades before he was first seen by Dr. Estes.1
In any event, whether or not Dr. Estes had reviewed the SMRs, it is important to bear in mind that his recitation of the veteran’s in-service diagnosis and treatment regimen was completely accurate, according to the pertinent SMRs. See R. at 10.
B. Court Precedent
The real effect of the Court’s holding here is that a “medical opinion” by a treating physician that his patient now suffers from a psychiatric problem that is connected to the patient’s service 20 years before is not evidence on the basis of which service connection could be awarded if the Board wished to do so, because the “medical opinion” derives from (accepting for the moment the majority’s incorrect factual inferences (see part II.A., above)) what the patient recounted to him about in-service treatment and symptoms. As I will attempt to illustrate below, that is not and should not be the law of this Court, and the cases cited by the majority certainly do not stand for that proposition.
1. Justus Case: The task for the Court in Justus, supra, as it is here, was to make a de novo determination of whether the appellant had presented “new and material” evidence to support reopening his claim. There, as here, the Board had held that a physician’s diagnosis was not “material” because it was “clearly based on the veteran’s medical history as reported by the veteran” and was “not based on a review of contemporaneous] medical records”. Justus, 3 Vet.App. at 512. The Justus Court, in holding that the physician’s statement was “clearly ‘new and material’ ” because it “contained information, not previously of record, which was relevant and probative of the issue at hand”, stated:
The error that the BVA committed was in assessing the credibility of the evidence prior to reopening the claim.
There is sometimes confusion in determining the ultimate credibility or weight to be accorded evidence, both questions of fact, and the issue of whether the evidence is new and material, a question of law. In determining the latter, the credibility of the evidence is to be presumed. This presumption is made only for the purpose of determining whether the case should be reopened. Once the evidence is found to be new and material and the case is reopened, the presump*481tion that it is credible and entitled to full weight no longer applies. In the adjudication that follows the reopening, the Board having accepted provisionally for reopening purposes the credibility of the new evidence, then must determine, as a question of fact, both the weight and credibility of the new evidence in the context of all the evidence, new and old.
Justus, 3 Vet.App. at 512-13 (internal citation omitted) (emphasis in original). The Court there concluded its analysis by noting that the presumption of credibility applied to the proffered evidence did not constitute an endorsement of “either the weight or the credibility” of that evidence, and that the weight and credibility determination “is a question of fact for the Board to decide upon reopening and readjudication with a statement of reasons or bases for its findings. See Gilbert v. Derwinski, 1 Vet.App. 49, 56-57 (199[0]).” Ibid, (emphasis added).
The Justus holding has been followed by no fewer than twelve subsequent prece-dential opinions of the Court2, and the majority’s conclusory statement that the Justus “presumption of credibility of the evidence does not come into play” in the instant case cannot withstand scrutiny in light of the facts of record.
2. Reonal Case: The majority attempts to distinguish the instant ease from Justus by relying on the Court’s recent holding in Reonal v. Brown, 5 Vet.App. 458 (1993). In Reonal, the Court held that the Justus presumption of credibility did not arise where the lack of in-service medical evidence of an in-service injury provided a plausible basis for the Board’s conclusion that a physician had based his opinions “solely on appellant’s account of his medical history” and where that history “had already been rejected ” in a prior final decision by an agency of original jurisdiction. Reonal, 5 Vet.App. at 460 (italics in original). In Reonal, the veteran’s SMRs were devoid of any in-service diagnosis of the claimed leg injury, while in the instant case the veteran’s SMRs clearly document the in-service diagnosis of a psychiatric disorder, “[hjysteria”. R. at 10. Thus, whereas in Reonal there was no objective independent basis upon which the physician could have based his statement, here the SMRs provide such a basis and, for the reasons stated in part II.A., above, I believe that the record on appeal strongly suggests that Dr. Estes relied on those records in forming his opinion. Again, at the very least, even if he was relying on what the veteran had related to him, the information about the veteran’s in-service medical history on which he relied was totally accurate, according to the SMRs.
Second, the record in this case shows that Dr. Estes had treated the veteran on numerous occasions over the course of several years before he wrote the August 1990 letter; he thus had ample opportunity to familiarize himself with the veteran’s history and to formulate his own independent “professional opinion” of the etiology of the veteran’s psychiatric disorder. In contrast, in Reonal there was no evidence that the physician in question had ever examined the appellant before the day that he wrote his statement. See Reonal, 5 Vet. App. at 460.
Third, in Reonal, as the Court there stressed, the factual allegations relied upon by the examining physician had, indeed, been “rejected” in a prior final regional office decision. Reonal, 5 Vet.App. at 461. Here, however, despite the majority’s assertion that the veteran’s “account of his medical history and service background ... have already been rejected by the RO and BVA”, ante at 478, the majority cites no record authority, and I can find none, in support of the proposition that the RO (in *4821967 or 1990) or the Board (in 1968 or 1990) had ever rejected the veteran’s recitations of, or the SMRs documenting, his in-service symptomatology. Indeed, it seems to me that the Board in 1968 conceded the in-service incident of “hysteria” (R. at 37, 38) and in 1990 conceded “nervous symptoms exhibited in service” (R. at 59).
Accordingly, the majority has failed to demonstrate how this case fits within the very narrow confines of the Reonal holding, and, therefore, the decision of the majority not to apply the Justus credibility presumption is unwarranted.
3.Hadsell Case: In its zeal to embrace Reonal, the majority overlooks this Court’s opinion in Hadsell v. Brown, 4 Vet.App. 208 (1993). There, the Board had stated in rejecting as not “material” a physician’s opinion as to the in-service etiology of the appellant’s disease: “ ‘[The physician’s statement] does not provide any thing [sic] more than the veteran’s or his attorney’s statements alleging a relationship between his current [disease]' and service’ ”. Hadsell, 4 Vet.App. at 209. The Court held that because the physician’s opinion “was, after all, a doctor’s statement that, based on reviewed facts, Crohn’s disease was present [in service]”, the Board “erred in its overly narrow reading of [the physician]^ statement as merely repeating the veteran’s earlier lay assertions”. Ibid. Further, the Hadsell Court stated as follows in holding that the Board erred in determining the credibility of the physician’s statement as part of the section 5108 step-one analysis: “We also note that the Board suggested that the reported history may not have been believed. However, a credibility determination is made properly only after reopening. See Justus, supra.” Ibid.
The facts in the instant case and Hadsell are alike in that the physician’s opinion in each case was the only medical evidence, old or new, that suggested a connection between the veteran’s current condition and his service. However, the facts of the instant case are far stronger as to the materiality of the medical-opinion evidence because here Dr. Estes was the treating physician whereas in Hadsell the physician had apparently never even seen the appellant; and because Dr. Estes stated his unequivocal opinion that the veteran’s current condition “is connected with conditions originating in the service”, R. at 62; Supp. R. at 2, whereas in Hadsell the physician’s advisory opinion stated only that such a connection was “certainly possible” and that the history related by the appellant’s attorney “suggested] a strong case of argument with the [VA].” Hadsell, 4 Vet.App. at 209.
4. Swann Case: As with its reliance on Reonal, the majority’s citation of Swann v. Brown, 5 Vet.App. 229 (1993), is groundless. See ante at 478. In Swann, the Court, in performing a section 5108 step-two review, was reviewing the BVA decision on the merits and found that the Board’s rejection of the physicians’ diagnoses of post-traumatic stress disorder was not clearly erroneous. That is a very different matter from a section 5108 step-one determination, as is required here, of whether or not such opinions would be new and material evidence, assuming (as is required by Justus, supra, when offered as a basis for reopening) their credibility.
5. Guerrieri Case: Finally, the majority’s analysis overlooks the fact, noted in part II.B.3, above, that Dr. Estes’ August 1990 statement was the only medical evidence in the record linking a current disability to the'veteran’s service. In attempting to reject Dr. Estes’ August 1990 medical opinion on the basis of its factual predicate, the majority also fails to take cognizance of the Court’s analysis in Guerrieri v. Brown, 4 Vet.App. 467 (1993). There, in determining that two physicians’ statements were “new and material” even though those statements offered a medical conclusion that had already been rejected by the Board in a prior final decision, the Court found that “the additional examination and medical skill underlying the physicians’] opinions brings new significance to the sum total of evidence” and stated as follows: “The probative value of medical opinion evidence is based on the medical *483expert’s personal examination of the patient, the physician’s knowledge and skill in analyzing the data, and the medical conclusion that the physician reaches.” Guerrieri, 4 Vet.App. at 470-71. In the present case, Dr. Estes’ statement is not a recitation of alleged facts, but instead constitutes his “medical opinion” based on his several years of experience in treating the veteran. Accordingly, because Dr. Estes’ numerous examinations of the veteran and his medical skill necessarily undergird his August 1990 letter, that letter “brings new significance to the sum total of evidence”, Guerrieri, supra, and is therefore “new and material” under the Guerrieri analysis.
C. Principles of Sound Medical Practice
Not only is the majority’s opinion inconsistent with the record and Court precedent, but it contradicts accepted principles of medical practice. Cf. Guerrieri, supra. The logical extension of the majority’s holding would appear to be that only a diagnosis untainted by a physician-patient dialogue could serve as a basis for reopening the veteran’s claim. Yet, the practice of taking and evaluating a patient history is an integral part of good medical care and certainly is vital to the successful treatment of mental disorders. Indeed, the section of VA’s Physician’s Guide For Disability Evaluation Examinations pertaining to psychiatric examinations states that “skill in interviewing ... [is] essential to an examination”, and sets forth a suggested “psychiatric interview technique”, in which the examiner is instructed to, among other things, “ask[ ] the veteran to relate everything that is troubling him/her”, “[pjrobe each symptom and how long it is present”, and inquire: “Has the veteran been thoroughly examined previously? What has he/she been told?” Physician’s Guide FoR Disability Evaluation Examinations § 14.10 (1985). Furthermore, “[t]he eliciting of information through the interview, plus an interpretation of the material contained in the claim file, an abstract of such data, [or] the medical record file ... should furnish the examiner with sufficient facts to be used as a basis for ... [a] diagnosis”. Ibid, (italics in original).
III. Conclusion
In essence, then, because the facts of this case do not support the majority’s assertion that the Board in a prior decision had rejected the veteran’s contentions about his in-service experiences, the majority is actually determining on the first step of the requisite section 5108 analysis that the Board was not clearly erroneous in rejecting a treating physician’s opinion that the veteran’s “problem with psychosis ... is connected with problems originating while in service.” R. at 62; Supp. R. at 2.
If, as I believe, the majority’s attempt to shoehorn the framework of the section 5108 step-two analysis into the first step stems from a basic disagreement with the Justus presumption of credibility, the proper forum to change that rule would be the Court en banc, not a panel opinion. See Bethea v. Derwinski, 2 Vet.App. 252, 254 (1992) (“Where there is an earlier panel or en banc opinion, ... a panel or single judge may not render a decision which conflicts materially with such earlier panel or en banc opinion.... Only the en banc Court may overturn a panel decision”).
A forthright approach to establishment of Court precedent would require the majority either to explain how this case is distinguishable from those cases in which the Court has applied the Justus credibility presumption, especially given the facts of Justus, supra, itself, and Hadsell, supra, or to send this case to the en banc Court for reconsideration of Justus. Instead, the majority is unnecessarily blurring this Court’s new-and-material evidence jurisprudence. When the facts of and analyses in this case and in Justus, Hadsell, Swann, and Guerrieri are examined, I do not envy those assigned the task of synthesizing our case law in order to provide comprehensive guidelines to VA adjudicators.
. The majority’s faith in the accuracy of the veteran’s powers of recall is apparently not undermined by his January 28, 1992, motion and affidavit to appeal without costs, where, asked to state the last date he had been employed, the veteran responded: “I am mentally ill and cannot remember the exact date.”
. Camphor v. Brown, 5 Vet.App. 514, 518 (1993); Person v. Brown, 5 Vet.App. 449, 450 (1993); Fluker v. Brown, 5 Vet.App. 296, 298 (1993); Sklar v. Brown, 5 Vet.App. 140, 145 (1993); Cox v. Brown, 5 Vet.App. 95, 98 (1993); Goodsell v. Brown, 5 Vet.App. 36, 41 (1993); Bernard v. Brown, 4 Vet.App. 384, 389 (1993); Hadsell v. Brown, 4 Vet.App. 208, 209 (1993); Chipego v. Brown, 4 Vet.App. 102, 105 (1993); Triplette v. Principi, 4 Vet.App. 45, 49 (1993); Rowell v. Principi, 4 Vet.App. 9, 18 (1993); Cuevas v. Principi, 3 Vet.App. 542, 547 (1992).