A large number of studies suggest that individual characteristics of psychotherapists affect the outcome of psychosocial treatments for psychiatric illness, but little work has been done to see if this is also the case for pharmacotherapy. In the context of a multicenter study that compared psychosocial and medication treatments for panic disorder, we assessed whether such characteristics as age of the psychiatrist, number of years of experience, and gender influence the outcome of treatment with the antipanic drug imipramine. Data were examined by multiple and logistic regression analyses for eight psychiatrists who treated a total of 57 patients with panic disorder. More physician experience, measured as years since completing residency, was associated with better response to imipramine on one of two main dichotomous measures (the Clinical Global Impression-Improvement Scale) and on six of nine continuous measure rating scales. Associations between psychiatrist age and outcome and between psychiatrist gender and outcome were also present but on fewer measures. Although these are post-hoc analyses that were not planned when the multicenter study was originally designed and therefore there are limitations in the information available about the psychiatrists' characteristics, the findings suggest that even in the context of a clinical trial that employs a specific protocol and single medication, physician experience may influence patient outcome. Depression and Anxiety 17:88-93, 2003.