Increasing doses of the benzodiazepine, diazepam, or placebo were administered to normal volunteers, drug-free alcoholic inpatients or patients with premenstrual syndrome who were participating in a clinical trial of the benzodiazepine alprazolam in a double-blind, cross-over study. In one series of experiments, subjects were pretreated with either placebo or the benzodiazepine antagonist, Ro-15-1788. In another series, subjects were pretreated with placebo or a high or low dose of caffeine. Following each dose of drug saccadic eye velocity, diazepam blood levels, plasma cortisol, and growth hormone were measured and self-ratings of anxiety and sedation were performed. After every other dose cognitive testing of memory and attention was performed. The effects of diazepam on these variables was quantified and diazepam dose response curves constructed. These dose response curves provide a measure of benzodiazepine receptor sensitivity in humans, as well as an evaluation of the ability of specific and nonspecific antagonists to block the actions of benzodiazepines. The studies involving patients participating the clinical trial of alprazolam will allow an examination of the effects of acute and chronic benzodiazepine treatment on receptor sensitivity.