Two putative predictors of placebo response were studied in three samples of psychiatric outpatients. Two groups, 73 university medical center patients and 56 college health service patients, underwent 1 week of placebo treatment. A quasi-control group of 112 patients receiving no medication waited about 1 week before beginning psychotherapy. One attribute, acquiescence or traditionalism, predicted placebo response, thereby replicating prior findings. Acquiescence was unrelated to change in controls indicating that it may represent a correlate of true placebo effect under some conditions. Additional findings suggested qualifications as to the generality of the relationship. The second attribute, autonomic awareness, was associated with change in all three samples. It appeared to predict nonspecific improvement unrelated to placebo probably due to its relationship to intensity of somatization.