Two hundred six outpatients with panic disorder and agoraphobia were randomly assigned to receive 4 weeks of treatment with placebo or sustained-release adinazolam under double-blind conditions. Eighty-eight percent of patients receiving drug and 85% of patients receiving placebo remained in the study at week 4. This report describes the "intent-to-treat" analysis of 202 patients who made at least one follow-up visit after randomization at baseline. On the basis of the Clinical Global Impressions-Improvement Scale, 69.7% of the adinazolam-treated patients were much or very much improved compared with 39.6% of the placebo-treated patients at week 4 or end-point (p = 0.0001). At week 4, panic attacks were completely blocked in 57.1% of adinazolam-treated patients and in 39.2% of the placebo-treated patients (p = 0.009). Adinazolam sustained-release treatment was statistically more effective than placebo treatment on measures of global improvement, number of panic attacks, SCL-90 phobia severity, main phobia severity, and anticipatory and general anxiety. No drug-placebo differences were found for overall self-rated phobia severity, unexpected or situational panic attacks, or for work, family, or social disability.