On June 1, 1990, the Board of Veterans’ Appeals (BVA) denied appellant’s claim for an increased rating for his 30% service-connected post-traumatic stress disorder (PTSD). Because the BVA failed adequately to consider and analyze the evidence presented and to explain its refusal to upgrade appellant’s rating, its decision is vacated and the case remanded for proceedings consistent with this opinion.
I.
Factual Background
Appellant served on active duty with the United States Army as a medical technician *537(medic) from 1947 to 1953. R. at 50. During part of this period, he was stationed in Korea at the height of the Korean conflict and was exposed to several traumatic combat experiences. R. at 8, 11, 50. During one such experience, he had a mental breakdown and was evacuated. R. at 11.
Subsequent to his honorable discharge in 1953, appellant was hospitalized several times because of his mental condition, including suicide attempts. R. at 8, 22, 63. In 1963, the Veterans’ Administration (now Department of Veterans Affairs) (VA) adjudicated his nervous disorder as service connected. R. at 50.
From September 14, 1988, to October 14, 1988, he was hospitalized by the VA for PTSD, agitated depression, and suicidal ideation. R. at 8-11. On September 23, 1988, the VA apparently reopened appellant’s past claim for nervous disorder and awarded him a 10% service-connected disability rating for anxiety reaction with PTSD. R. at 11, 12.
In November 1988, appellant was examined by VA psychiatrist Dr. Jose Amato, who diagnosed appellant as having chronic PTSD along with recurrent major depression. Dr. Amato gave a “guarded prognosis” for appellant and indicated that his condition could be expected to last twelve months or more. R. at 16. This diagnosis was confirmed on February 13, 1989, by VA Dr. Dennis Grant who treated appellant for long-term sleep disorder with nightmares from his war experiences. R. at 14-15. However, in evaluating appellant, Dr. Grant also noted:
[T]he ... patient [was] ... alert, cooperative, and in good contact with reali-ty_ His speech was ... coherent, [his] mood was labile_ He was somewhat anxious ... [with no] evidence of suicidal thoughts.... [He showed] no evidence of psychotic behavior.... His memory[,].... judgment and reasoning appeared intact. He did show depressive affect, however.
[An increased dosage of antidepressant during hospitalization] resulted in a dramatic improvement in his depressive affect. This improvement was maintained thereafter.... He demonstrated a very high and positive motivation and productive involvement in the [PTSD] program. He improved in social isolation.... However, there was a tendency toward developing recurrent depressive affect. ...
His employability is somewhat limited in a competitive labor market at the present time_ Employability could be improved if vocational assistance and retraining could be done on an outpatient basis.
R. at 14, 15.
On November 30, 1988, appellant filed a disability claim for an increase in his PTSD rating which was denied by the VA Regional Office (RO) on March 10, 1989. R. at 17-18, 20-21.
Shortly after this denial, on March 11, 1989, appellant underwent another VA examination. This time he was treated by Dr. Marcelle Leet who reconfirmed the diagnosis of PTSD and observed:
The [appellant] ... feels estranged_ tends to avoid crowds and is distrustful.... [He] tends to avoid activities and events which remind him of his expe-riences_ [He] is frequently hypervi-gilant, startles easily, and suffers from chronic sleep disturbance.
[Appellant’s] thought processing is logical and goal directed. There is no evidence of psychotic thought content.... [or] suicidal ... ideation. He appears to have adequate recent and remote memory as well as concentrating ability....
R. at 22, 23. In profiling appellant’s history, Dr. Leet reported:
The [appellant] attended school through the eighth grade and received a GED. He also has had three years of college with specialized training in surgical technology. [Appellant] has worked as a surgical technologist. He last worked in 1978 in this field. He is unable to obtain employment in this area owing to *538his felony conviction. As a result he has worked at a number of occupations including an administrative clerk, truck driver, farm laborer, etc. ... [Appellant] states that, although he is physically capable of working, he has ... frequently ... walked off jobs due to irritability and fear.
Id. (emphasis added). Dr. Leet also recorded:
[Appellant] thinks that [his mental and emotional condition] has had a number of ruinous affects [sic] on his life, including the fact that he spent 14 years in prison for forgery. He states that because of his constant stress and anxiety he basically ran away and could not support himself, thus being forced to forge a check. [Appellant] has been hospitalized on a number of occasions owing to fear and suicide attempts....
R. at 22.
On May 18, 1989, the RO, in apparent response to an October 31, 1988, claim filed by appellant, issued a decision in which it raised appellant’s PTSD rating to 30%. R. at 31-34, 38. In addressing appellant’s em-ployability, it stated:
[Appellant’s] employability was felt to be somewhat limited in a competitive labor market because of both his medical and psychological difficulties. It was felt that employability could be improved if vocational assistance and retraining could be done on an outpatient basis.
Id. at 33.
Dissatisfied with this rating, appellant filed a Notice of Disagreement (NOD) with the RO on June 27, 1989. R. at 39. In support of his claim, appellant submitted the results of an independent comprehensive psychological examination performed June 29, 1989, by Dr. George DeLong, private psychologist. After conducting a series of diagnostic tests on appellant, Dr. DeLong stated that appellant’s manifestations included exaggerated startle response, periods of pressured and rapid speech with elevated heart rate and breathing when attempting to discuss combat experiences, hypervigilance, scanning, flashbacks, and nightmares, and that he suffered from severe chronic PTSD. R. at 43.
In one of his reports dated July 5, 1989, he classified appellant’s impairment as follows: mild impairment in personal habits; moderate impairment in the ability to socialize with friends and neighbors, attend meetings, work around the house, perform simple tasks, and understand, carry out, and remember instructions; moderately severe impairment in the ability to relate to other people, maintain outside interests, respond appropriately to supervision and coworkers, and perform complex and repetitive tasks; and severe impairment in the ability to respond to customary work pressures and perform varied tasks in a routine work setting. R. at 45-46. With respect to appellant’s employability, Dr. Delong further opined that he did not believe that appellant was “capable of completing ... a [vocational assistance and retraining] program and entering into competitive employment ... [as h]is symptoms ... [are] quite severe ... and will remain so for the foreseeable future.” R. at 43-44.
On November 15, 1989, the RO, after considering this evidence, denied appellant’s claim for an increased PTSD rating. R. at 55-57. Appellant filed an NOD with this rating decision and appealed to the BVA on December 12, 1989. R. at 58-59. In his appeal, appellant sought a higher rating and a remand to the RO for “complete psychoneurological examination due to the fact that the last examination was given March 11, 1989.” R. at 69.
On June 1, 1990, the BVA denied appellant’s request for another examination and affirmed the RO decision, finding that the
veteran remains oriented and goal-directed, with logical thought processes, and adequate memory and concentration abilities, such that social and industrial impairment is not more than definite [and thus] ... not more than 30 percent disabling. ...
William E. Cousino, BVA 90-, at 6 (June 1, 1990). Appellant subsequently appealed to this Court.
II.
Analysis
PTSD is rated under 38 C.F.R. § 4.132, Diagnostic Code 9411 (1991) (DC 9411) which provides for the following cate*539gories of disability: 100% where there is “[t]otally incapacitating psycho-neurotic, symptoms bordering on gross repudiation of reality” and there is demonstrable inability “to obtain or retain employment”; 70% where there is severe impairment in the “[a]bility to establish and maintain ... relationships” and “there is severe impairment in the ability to obtain or retain employment”; 50% where there is considerable impairment in the “[ajbility to establish and maintain ... relationships” and there is “considerable industrial impairment”; and 30% where there is “[d]efinite impairment in the ability to establish or maintain ... relationships” and there is “definite industrial impairment”. While social impairment is to be considered in determining disability, it is only relevant to the extent of its impact on the veteran’s degree of industrial impairment. See Webster v. Derwinski, 1 Vet.App. 155, 158-59 (1991); 38 C.F.R. § 4.129 (1991). Thus, these categories determine disability based solely on a veteran’s actual industrial impairment.
Further, in evaluating PTSD, both 38 C.F.R. § 4.130 (1991) and 38 C.F.R. § 4.2 (1991) are relevant. Section 4.130 states:
[T]he objective findings and the examiner’s analysis of the symptomatology are the essentials. The examiner’s classification of the disease as “mild,” “moderate,” or “severe” is not determinative of the degree of disability, but the report and the analysis of the symptomatology and the full consideration of the whole history by the rating agency will be.
and section 4.2 instructs:
It is the responsibility of the rating specialist to interpret reports of examination in the light of the whole recorded history, reconciling the various reports into a consistent picture so that the current rating may accurately reflect the elements of disability presented....
While purporting to review all evidence in accordance with the above regulatory provisions and determining that a 30% rating was appropriate, the BVA stated:
While it is apparent that the veteran does experience psychiatric symptoms of some significance, with anxiety and depression, as well as nightmares and sleeping difficulties, and an isolative pattern of existence being amongst those problems characteristic of his disorder, it should be pointed out, nevertheless, that the 30 percent rating currently in effect contemplates a degree of impairment of definite proportions. In our judgment, the extent and the severity of the indicated symptoms, viewed in terms of their impact upon factors which would affect his employment, such as his initiative, flexibility, and reliability, are found to be commensurate with, or accurately reflected by, the 30 percent evaluation presently in effect.
In reaching our determination, the Board has extended consideration to the veteran’s entire psychiatric disability picture, but has found no perceptible basis for supporting a finding that pathology which can reasonably be attributed to a stress-related disorder is productive of more than definite social and industrial impairment.
Cousino, BVA 90-021295, at 5-6 (emphasis added).
Despite this analysis, there are unresolved evidentiary conflicts in the record on the degree of appellant’s industrial and social impairment. Dr. Grant’s February 1989 report states that appellant’s mental condition dramatically improved as a result of medications and participation in the VA PTSD program and that appellant’s em-ployability, although somewhat limited, “could be improved if vocational assistance and retraining could be done on an outpatient basis.” To the contrary, Dr. De-Long’s June 1989 report states that appellant’s PTSD symptoms (i.e., exaggerated startle response, periods of pressured and rapid speech with elevated heart rate and breathing when attempting to discuss combat experiences, hypervigilance, scanning, flashbacks, and nightmares) were quite severe, resulted in appellant’s social dysfunction, and prevented him from completing a vocational rehabilitation program and being gainfully employed. Thus, the BVA selectively referenced only some of appellant’s symptoms without addressing or reconciling the report of Dr. DeLong which appears to indicate a disability in excess of *54030%. See R. at 14-15, 43-44. See also R. at 8-11, 22-23, 40-42, 45-46. Indeed, since the RO had raised appellant’s PTSD rating to 30% before submission of Dr. DeLong’s report and the BVA affirmed the RO’s determination, it appears that neither the RO nor the BVA ever accorded any weight to his report.
Moreover, the BVA failed to discuss Dr. Leet’s observations regarding appellant’s frequent hypervigiliance, chronic sleep disturbance, startle response, and social estrangement as they relate to appellant’s degree of disability. Finally, it failed to deal with the effect that appellant’s felony conviction and incarceration, referenced by Dr. Leet, have on his employability and to what extent, if at all, this criminal activity was attributable to his service-connected condition. The BVA is required to address thoroughly, analyze carefully, and reconcile all relevant evidence in the record. See, e.g., Washington v. Derwinski, 1 Vet.App. 459, 463-64, 466 (1991); Willis v. Derwinski, 1 Vet.App. 66, 68-70 (1991); Fletcher v. Derwinski, 1 Vet.App. 394, 397 (1991); Hatlestad v. Derwinski, 1 Vet.App. 164 (1991); Webster, 1 Vet.App. at 158-59; Ohland v. Derwinski, 1 Vet.App. 147, 149-50 (1991); Wilson v. Derwinski, 1 Vet.App. 139, 139-40 (1991); Gilbert v. Derwinski, 1 Vet.App. 49, 56-57 (1990). It has failed to do so.
The Court notes that because there has been a wide diversity of medical opinions on appellant’s condition, the BVA’s denial of appellant’s request for a complete psychoneurological examination was viola-tive of the duty-to-assist requirement in 38 U.S.C. § 5107(a) and, hence, arbitrary and capricious. 38 U.S.C. §§ 5107(a) (formerly § 3007(a)), 7241(a)(3) (formerly § 4061(a)(3)). See, e.g., EF v. Derwinski, 1 Vet.App. 324, 325-26 (1991); Green v. Derwinski, 1 Vet.App. 121, 123-24 (1991); Akles v. Derwinski, 1 Vet.App. 118, 120-21 (1991). Consequently, a medical examination should be performed, pursuant to 38 C.F.R. § 3.327(a) (1991), to assess appellant’s service-connected disorder. See Akles, at 121. Because of the circumstances presented here, the Court directs that this examination be performed by a physician or physicians who have not previously examined appellant and that such examiners be provided with all VA medical records relating to appellant’s PTSD. In addition, if deemed necessary for an accurate assessment of appellant’s condition, an independent medical opinion may also be obtained under 38 C.F.R. § 3.328(a) (1991). The results of such examination, and the opinion, if obtained, are to be provided to appellant who is also free to submit additional relevant evidence in support of his claim. Cf. Lichtenfels v. Derwinski, 1 Vet.App. 484, 488 (1991); Colvin v. Derwinski, 1 Vet.App. 171, 175 (1991).
III.
Conclusion
For the reasons stated above, the decision of the BVA is vacated and remanded for proceedings consistent with this opinion.
It is so Ordered.