Mixed-panic disorder patients (16/60, 27%) randomly assigned to receive blind placebo during a 40-week treatment study were defined as placebo responders based on combined criteria of Hamilton Anxiety Scale score percentage decreases below the median point (-42%), moderate to marked improvement on both clinician and patient Clinical Global Impressions scores, and panic-free at final treatment visit. These criteria applied separately also resulted in a similar clinical grouping and pattern of response. Differential patterns of response between responders and nonresponders occurred across most clinical measures of panic/anxiety. Responders experienced early improvement within the first week of double-blind treatment. This response progressed during treatment and tended to persist during taper and at followup 1 month later. Post-hoc analysis of demographic and clinical features at entrance into the study failed to characterize this stringently defined group of placebo responders.