The efficacy and safety of buspirone (Buspar), diazepam (Valium), and placebo were compared in a 3-week double-blind study of 131 patients with scores of at least 15 on the Hamilton Rating Scale for Anxiety. Active drugs were administered orally in 5-mg tablets to a maximum dose of 10 mg t.i.d. Buspirone was given to 43 patients, diazepam to 46 and placebo to 42. There was significant improvement in all three treatment groups at 1 and 2 weeks, with further significant improvement with both active drugs, but not with placebo, at 3 weeks. The incidence of side effects reported voluntarily by patients was similar with the two active drugs, but drowsiness was significantly more frequent with diazepam.