The lifetime rate of suicide attempts among subjects in the Epidemiologic Catchment Area database with bipolar disorder, unipolar disorder, and any other DSM-III-defined Axis I disorder were determined. The latter constituted a control or reference group. The lifetime rates of suicide attempts of persons in these diagnostic groups were 29.2%, 15.9%, and 4.2%, respectively. Multivariate logistic regression analysis was used to calculate the odds ratio of subjects within a diagnostic group having a history of a suicide attempt relative to subjects in a second group. The odds ratio of subjects with bipolar disorder having a history of a suicide attempt relative to subjects in the control group was 6.2 (df 1, x(2) = 5347.2, p < .0001). The odds ratio of subjects with unipolar disorder having a history of a suicide attempt relative to subjects in the control group was 3.1 (df = 1, x(2) = 4785.2, p < .0001). The odds ratio of subjects with bipolar disorder having a history of a suicide attempt relative to unipolar subjects was 2.0 (df = 1, x(2) = 697.9, p < .0001). Bipolar disorder has a strong relationship to a history of suicide attempts relative to unipolar disorder and other Axis I disorders.