Social skills role-play assessment has been plagued by a lack of external validity. One solution recommended has been to select role-plays that are personally relevant for clients. The present study assessed the relationship between the personal relevance of role-plays and degree of socially skilled behavior. Results failed to support the hypothesis that clients would perform most poorly in those role-plays they viewed as most personally relevant. Results suggest that it may not be critical for therapists to develop role-plays that are personally relevant; the variables controlling socially skilled behavior do not seem unique to personally relevant situations. Instead, it may be more important for therapists to examine a broad range of social situations. If only personally relevant social skill situations are assessed, an incomplete picture of a client's range of social deficits may be obtained.