To evaluate the efficacy of alprazolam in the treatment of premenstrual syndrome. A randomized, double-blind, placebo-controlled, crossover trial of alprazolam during eight menstrual cycles. Outpatient clinic at the National Institute of Mental Health, Bethesda, Md. Twenty-two women with prospectively confirmed premenstrual syndrome entered this study. All subjects were either self-referred or were referred by their physicians. All reported having regular menstrual cycle lengths, were taking no medication, and were free of current or recent medical or psychiatric illness. Two subjects did not complete the trial. Participants were assigned to receive alprazolam or placebo as follows: cycle 1, 0.25 mg of alprazolam or placebo three times daily beginning on menstrual cycle day 16; cycle 2, 0.50 mg of alprazolam or placebo three times daily according to the regimen during the first cycle; cycles 3 and 4, 0.75 mg of alprazolam or placebo three times daily from menstrual cycle day 16 and continued throughout the fourth menstrual cycle to evaluate the efficacy of relatively long-term (approximately 6 weeks) treatment with alprazolam. Daily self-report symptoms ratings were completed during the entire study period. We observed no significant differences in the severity of premenstrual symptom ratings during alprazolam administration compared with placebo on any scale except the Beck Depression Inventory Scale. The Beck Depression Inventory ratings demonstrated a statistically (F1,19 = 7.1, P < .05), but not clinically, significant improvement in depressive symptoms during alprazolam administration compared with placebo. Our findings do not support alprazolam as a uniformly effective treatment for the symptoms of premenstrual syndrome.