Theorizing by Bakan (1966), Bales (1955; 1958), Block (1973), and others suggests that interactional involvement may require the application and synthesis of both traditionally "masculine " (i. e., instrumental) and traditionally "feminine " (i. e., expressive) capacities. This view is supported not only by a previous study of sex role influences in mixed-sex dyads (Ickes and Barnes, 1978) but also by the present study of sex-role influences in same-sex dyads. The degree of interactional involvement (e.g., talking, looking, gesturing) was found to be high for dyad types in which both instrumental and expressive capacities could be integratively applied, but was found to be low for dyad types in which only one set of capacities (instrumental or expressive) could be represented. The degree of satisfaction that dyad members expressed regarding their interacctions appeared to vary according to the degree to which the level of interactional involvement they experienced was consistent or inconsistent with their own predisposition to be expressive. These findings are discussed in terms of their implications for current conceptualizations of sex-role identification.