This paper presents two sets of data that suggests a weak but specific placebo response in agoraphobia. First, analyses in 20 agoraphobic patients given single-blind placebo over a 2-week period, without the customary confound of instructions for exposure to phobic situations, revealed a statistically significant reduction in panic and phobic symptoms. However, symptoms remained in the moderate to severe range and functioning was virtually unchanged. Second, comparisons between six agoraphobic patients receiving double-blind placebo and six others receiving "no pills," matched for age, sex and exposure treatment, revealed a significant placebo effect over an 8-week period. Implications for clinical research are briefly discussed.