Racial variations in the use of effective medical care and subsequent clinical outcomes have been identified for many medical conditions. Still, it is unclear whether racial variations in care and clinical outcomes exist for depressed primary care patients. Primary care patients presenting for routine treatment were screened for major depression as part of a study to disseminate a depression treatment guideline. Primary care physicians (PCPs) were informed of their patients' depression via an electronic medical record system and asked whether they agreed with the diagnosis. Treatment patterns and depressive symptoms over the following six-months were assessed by chart review and the Hamilton Rating Scale for Depression, respectively. Over a 20-month period, 8,944 African-American and Caucasian patients aged 18-64 were approached for screening. African-Americans were less likely to agree to undergo screening than Caucasians (83% vs. 88%; P<.0001), but those doing so were more likely to report mood symptoms (26% vs. 15%; P<.001). 204 patients, including 52 African-Americans (25%), met protocol-eligibility criteria and completed a baseline interview. Baseline sociodemographic and clinical characteristics, and PCPs' agreement rate with the depression diagnosis were similar. Although PCPs were less likely to counsel their African-American than Caucasian patients for depression (P=.03), this difference resolved after adjusting for education level, employment, and insurance status and we found no other variations in the depression care provided or in clinical outcomes by race. We found little racial variation in either process measures or clinical outcomes for depression in our sample of African-American and Caucasian primary care patients.