Psychiatric patients who were classified as black, white or Hispanic at the admitting office, were divided into broad diagnostic groups according to the psychiatrist's diagnosis at termination. Chi-square analyses comparing ethnicity and diagnosis yielded no significant differences. A psychologist administered the Structured Clinical Interview in the admitting office prior to determination of precipitating events and assignment of diagnosis. When scores were compared by analysis of variance, no significant group differences were found on over-all severity of illness, but 3 of the 10 subareas showed significant differences. The results may reflect cultural and social influences.