Schizophrenia is diagnosed more frequently among African Americans while mood disorders are identified more often among whites. Such findings have raised serious questions about the accuracy of clinical judgment. This article analyzes data on 665 African American and white psychiatric inpatients using a semi-structured diagnostic instrument. The paper explores the relationship of patient race to schizophrenia, schizoaffective disorder, major depression, and bipolar disorder. The paper also explores the extent to which patient race is related to the manner in which clinicians link individual symptoms to diagnoses. Results indicate some significant race differences in diagnosis remain even when a semi-structured instrument and DSM criteria are used, whites, were more likely than African Americans to receive a diagnosis of bipolar disorder and less likely to be diagnosed with schizophrenia. There were no race differences in major depression. Some patterns of symptom attribution differed by race. The results are consistent with previous sociological research showing that patient race is related to diagnosis even when standardized diagnostic criteria are used. These findings underscore the importance of clinical judgment within the context of cross-race and cross-ethnic diagnosis. Clinical training programs must reduce ethnocentric bias by teaching the appropriate use of the socio-cultural information necessary to employ DSM-IV's Cultural Formulation.