Buspirone, a non-sedating anxiolytic, has yielded contradictory results in smoking cessation pilot studies and trials. We tested buspirone (n = 51) versus placebo (n = 49) in a placebo-controlled, double-blind trial of smoking cessation. Survival analyses were performed with use of strict abstinence criteria for efficacy (carbon monoxide levels < or = 8 ppm; no self-reported slips to smoking). No treatment differences were observed between active and placebo groups. There were also no differences among "anxiety" level groups formed post hoc from high versus low, pre-quit anxiety test scores. A number of withdrawal symptoms increased significantly after subjects quit smoking for both the active drug and placebo groups, but these symptoms were not relieved by treatment. There appears to be little evidence that buspirone is effective in smoking cessation or in the relief of withdrawal associated with cessation in a general sample. Selecting for generalized anxiety or anxiety related to cessation is suggested for future testing.