Thirty-two outpatients with a DSM-III diagnosis of panic disorder or agoraphobia with panic attacks were randomly assigned to 4 weeks of treatment with clonazepam or placebo, after a 1-week placebo washout period. Twenty-nine patients entered the double-blind phase of the study and were eligible for intent-to-treat analysis. Clonazepam-treated patients experienced significantly fewer panic attacks, and these were of lesser intensity and short duration than those in placebo-treated patients (p < 0.001). Clonazepam was also superior to placebo with respect to symptoms of anxiety and depression (p < 0.001). The mean dose of clonazepam at week 4 was 2.2 mg (standard deviation, 0.7 mg). There was significant (p < 0.05) correlation between drug concentration in plasma and decrease in the global measure of the severity of panic disorder (r = 0.68); similar trends were seen for the decreases in frequency (r = 0.60) and severity (r = 0.55) of panic attacks, but not between concentration in plasma and decline in generalized anxiety. The most common adverse event was drowsiness, which occurred in 9 of 13 clonazepam-treated patients.