The relationship between self-monitoring and boundary spanning was investigated in this study. In an attempt to isolate the behaviors used by self-monitors to facilitate performance in a boundary spanning role, a laboratory setting was used to videotape subjects assuming a contrived boundary spanning role. Subjects (N=234) were introductory psychology college students who were divided into high, middle, and low self-monitoring groups based on scores on the Snyder Self-Monitoring Scale. After viewing two videotaped exchanges between a model "Campus Escort" and "visitor," subjects were shown short video sequences of a visitor expressing his interest and concerns. Following each sequence, subjects role played their response to the visitor as a Campus Escort (role play condition) or selected the most appropriate responses (multiple choice condition). No differences were found regarding the appropriateness of the messages in either condition. High self-monitoring males used more response categories during their role plays than did any other group. High self-monitoring females were rated as more extraverted. An investigation of these results showed that for males, acting ability covaried with the number of response categories used during role plays, verbal fluency, and overall performance and with females' extraversion and friendliness ratings. These results suggest that the behaviors of the boundary spanners were not determined by their sensitivity to social cues but rather by their ability to assume a role and this role may be influenced by stereotypic gender differences. The ability to assume a role may be more important to boundary spanning roles than sensitivity and behavioral pliability to social cues. (ABL)