Although concrete behavior—such as avoidance, discrimination, rejection—is foundational to most definitions of stigma, knowledge of psychiatric stigma has been constructed mostly on the basis of measurement of self-reported attitudes, beliefs, and feelings. To help fill this gap, the current study examined avoidance behavior in psychiatric stigma. That is, we predicted that people would seek more physical distance from a man with a psychiatric problem than a man with a medical problem. One hundred fourteen undergraduates expected to meet a man with either Type II diabetes or schizophrenia. After completing several measures of self-reported stigma, participants eventually moved to an adjacent room and sat in one of several seats that systematically varied in their proximity to a seat ostensibly occupied by the target man. Results indicated that the expectation of meeting a man with schizophrenia, compared with diabetes, led to greater desired social distance, greater self-reported fear, and higher appraisals of the man’s dangerousness and unpredictability. More importantly, participants elected to sit farther away from the ostensible man with schizophrenia. This pattern of findings offers behavioral evidence of the psychiatric stigma phenomenon that has mostly been documented via measurement of self-reported attitudes and impressions. We hope that these results stimulate renewed interest in measuring stigma-relevant behavior in the laboratory setting.