We investigated whether self-reported racial/ethnic and gender discrimination were associated with lower rates of diabetes management. Analyses focused on 524 respondents with diabetes from the 2001 Commonwealth Health Care Quality Survey. Probit models estimated the marginal effect of self-reported racial/ethnic and gender discrimination on diabetes management, adjusting for racial/ethnic and gender concordance of the patient and physician, as well as a variety of health and sociodemographic indicators. Self-reported racial/ethnic discrimination was associated with a roughly 50% lower marginal probability of receiving a hemoglobin A1c test, foot exam, and blood pressure exam, when accounting for racial/ethnic and gender physician-patient concordance and other factors. Self-reported gender discrimination was associated with 22% lower marginal probability of conducting a hemoglobin test but not associated with the other management indicators. Evidence from this study suggests that racial/ethnic discrimination is an important barrier to diabetes management.