This study evaluates the effectiveness of standardized patients for teaching patient selection in esthetic surgery. Six actors received detailed character descriptions. Each was interviewed by a resident for 30 minutes in a conference setting. Participants completed a questionnaire, and a discussion was held. Written simulations were administered before and after the sessions, from which a faculty standard was developed using hierarchical cluster analysis. Resident responses were compared with the standard using a squared euclidean metric. Residents' pre- and posttest scores were compared with the faculty standard using 2-way analysis of variance. Accuracy scores were found to be significantly lower (more accurate) after the training than before (P < 0.001). Upon a 6-item questionnaire, both faculty and residents agreed that it was a worthwhile exercise (faculty mean, 6.2 out of 7; resident, 6.3) and that the standardized patients were believable. Standardized patients can provide effective instruction in traditionally difficult-to-teach areas such as communication and patient selection.