Despite the growing interest in the process of informed consent, little research has actually examined what information clinicians give to patients concerning psychiatric treatment. In this paper, the neuroleptic drug disclosure practices of 60 North Carolina psychiatrists are examined, utilizing survey data collected in 1979. Although the vast majority of psychiatrists surveyed reported that they routinely provided information to their patients regarding the reason for the chemotherapy, signs of drug toxicity, and minor side effects, information on major side effects and tardive dyskinesia was reported to be far less routinely supplied. Physician and patient/treatment context characteristics influencing the scope of physicians' drug information disclosure were examined through the use of bivariate correlation and multiple regression techniques. Patient racial and diagnostic characteristics, as well as physician medical and psychiatric journal use, were found to be significant predictors of drug disclosure in the multivariate analyses. Patient social class, patient legal status, and treatment setting were significant predictors of drug disclosure in the bivariate, but not in the multivariate analyses. Study limitations and implications are discussed.