This study attempted to address the performance inhibition hypothesis by assessing nonverbal social performance in socially anxious individuals during a task where verbal content was standardized, thereby decreasing the overall performance requirements, thus theoretically decreasing their social distress. Fifty-nine subjects were identified as high or low socially anxious and participated in two behavioral role-play tasks. Both role-plays included a standard heterosocial conversation task; however during the second task subjects were provided their verbal content through a bug-in-the-ear wireless transmitter. Results showed no significant within or between-group differences on measures of nonverbal social skill. However, a global rating of social skill revealed a significant group difference. These results do not support the performance inhibition hypothesis and support the notion that isolated behaviors aren't enough to distinguish socially anxious and non-socially anxious individuals from one another. Rather, it's the unique combination of all elements of social skill that allows for this differentiation.