In a multicenter, placebo-controlled, clinical trial, the efficacy of Limbitrol was compared with that of its components, amitriptyline and chlordiazepoxide. All patients had a diagnosis of primary depression. Data from 279 patients were evaluated using the Hamilton depression scale, the Beck depression inventory, and physician and patient global change measures. Statistically significant differences favoring Limbitrol occurred after 1 week of treatment, and a trend in favor of Limbitrol continued throughout the remaining 3 weeks. In most efficacy comparisons, the combination was as good as, or better than, amitriptyline alone. It was superior to chlordiazepoxide alone after 2 and 4 weeks of treatment. Each component produced an independent contribution to the total therapeutic effect: the chlordiazepoxide effect was more prominent in the first 2 weeks and the amitriptyline effect in the latter 2 weeks. A trend favoring amitriptyline over chlordiazepoxide was evident by week 4. The overall incidence of side effects was comparable in both Limbitrol- and amitriptyline-treated groups. Limbitrol-treated patients exhibited more sedation, but significantly fewer Limbitrol patients discontinued treatment prematurely because of side effects.