OBJECTIVE Published reports indicate that African Americans are underrepresented among patients treated with ECT. The reason for this disparity in practice has not been determined. This study addressed this question by using existing data on a large series of patients treated with ECT at a single academic medical center. METHODS The hospital's administrative databases were used to select Caucasian and African American patients with a diagnosis of major affective disorder treated over the period from November 1993 to March 2002. Independent variables were age, sex, treatment unit, readmission within 30 days, type of insurance, and geographic zone of residence. The dependent variable was likelihood of being treated with ECT, computed for each race group. RESULTS Caucasians were more likely than African Americans to be treated with ECT (odds ratio=4.71; 95% confidence interval [CI]=3.77-5.90). None of the variables examined provided an explanation for this disparity. When all of the variables were controlled simultaneously, the likelihood of being treated with ECT remained significantly higher for Caucasians than for African Americans (odds ratio=2.48; 95% CI=1.89-3.25). CONCLUSIONS The racial disparity in the use of ECT cannot be explained on the basis of the variables studied. It is not an artifact of the age of the patient population, nor can it be explained on the basis of insurance coverage, the social class of the patients, or their illnesses' being more treatment resistant. The authors propose several other hypotheses, including explanations relating to clinical presentation, differential response to other treatments, differences in patients' willingness to consent, and physicians' behavior, that could be explored by using other methods.