A study was conducted to test a model of relational competence that consisted of three components: motivation, knowledge, and skill in interaction for both participants in a conversation. The model is based on the assumption that the additive combination of these components for both participants will be significantly predictive of several competent outcomes, including communication satisfaction, perceived confirmation, conversational appropriateness, and effectiveness. Subjects, approximately 500 college students, completed measures of communication involvement, interpersonal communication apprehension (motivation), self-monitoring (knowledge), attentiveness, self-rated competence, and ratings of other competence (skill). The measures were completed before and after interaction in one of two dyadic situations--acquaintance and task. Results indicated that the context-independent measures did not contribute significant variance to the model. However, both the context-specific measures of skill contributed substantial amounts of variance in all of the predictions. The findings support the conclusion that competence is a contextual phenomenon and requires context-specific conceptualization and measurement. Appendices include statistical tables and a six-page list of references. (FL)