One hundred forty outpatients with major depression were admitted to an 8-week, placebo-controlled, double-blind study of buspirone. Entry criteria included a Hamilton Rating Scale for Depression (25-item [HAM-D]) score of greater than or equal to 18 and a Hamilton Rating Scale for Anxiety (HAM-A), score of greater than or equal to 18. A flexible dose schedule ranging from 5-90 mg/day was employed. The mean dose of buspirone was 41-54 mg/day from Week 2 to the end of the study. Sixty-four percent of buspirone patients and 50% of placebo patients were melancholic; 64% of buspirone patients and 74% of placebo patients discontinued treatment before the end of the study. Extender data analysis showed that buspirone patients had significant (p less than .05) HAM-D score reductions compared with the placebo group at Weeks 2, 3, 4, and 6. The HAM-D retardation factor trended toward significance over placebo at Weeks 3, 4, and 6. HAM-D change scores for the subgroup of melancholic patients taking buspirone were significantly (p less than .02) better than those of the placebo-treated melancholic subjects at Weeks 2, 3, 4, and 6. Most other efficacy parameters also favored the buspirone-treated group over the placebo-treated group. The most common adverse experiences for the buspirone group were CNS effects (74% in the buspirone group vs. 21% in the placebo group) and gastrointestinal effects (55% in the buspirone group vs. 37% in the placebo group). Side effects consisted of dizziness, light-headedness, nausea, and headache. No serious or unexpected adverse effects occurred.(ABSTRACT TRUNCATED AT 250 WORDS)