The purpose of this study was to evaluate a drug history method. The method consisted of a review of systems (ROS) history format, which utilized common symptoms/diseases per organ system matched with corresponding drug treatments. Findings of this method were compared with drug histories documented in the medical record. The study was conducted over an eight-week period on general medicine inpatients and ICU patients. All data retrieved were categorized into total number of drugs found and then analyzed as to legend, nonlegend, and unidentifiable content. Findings were statistically analyzed using chi-square. During the study, 207 patients were admitted. Of these patients, 186 drug histories were obtained; 21 patients could not be interviewed. The ROS method documented 406 drugs as compared with 60 drugs by the medical record (p less than 0.001). Performance time was comparable to similar studies (14 +/- 5.85 min). Using a concise, logically organized technique, more drugs were identified than reported by the medical record. We conclude that the experimental method is a productive method for obtaining a drug history.