Persons with schizophrenia commonly have impaired social functioning (Wallace 1984). Those with greater impairments, particularly as measured by premorbid social attainment, have a poorer clinical prognosis (Strauss and Carpenter 1972). Interventions designed to improve social competence, such as social skills training, have yielded generalizable and durable effects and may have reduced relapse rates (Wallace and Liberman 1985; Liberman et al. 1986; Hogarty et al. 1986). Thus, a valid measure of social skills should be a useful clinical and research tool. This research explores the validity of a new instrument for measuring such skills, the Assessment of Interpersonal Problem Solving Skills (AIPSS) (Donahoe et al., this issue). The AIPSS differs from more conventional social functioning measures because it 1) utilizes observations of role-playing, rather than self-report or third-party report; 2) provides a rating of the patient's current rather than past functioning; 3) involves videotaped simulated "real life" situations that pose challenges to the patient's ability to solve socially relevant problems; 4) permits assessment of patient's social perception, processing of social information for action planning, and verbal and nonverbal social responses. In this study, we examined the relationship between these parameters of schizophrenic patients' social functioning and their recent relapse history.