To compare the efficacy of five versus one session of Behavioral Counseling in a 12-week intervention to increase self-efficacy and family and friend support for activity, and examine self-efficacy and support as mediators of activity among 46 urban women. A randomized, controlled trial conducted during 2004 in Omaha, Nebraska. Outcomes were analyzed with Repeated Measures-ANOVA and path analysis. No significant change was observed in self-efficacy in the five-session group, but a significant decrease was observed in the one-session group (p = .005). Family and friend support increased significantly in the five-session group (p < .001, p = .019). The intervention effect on activity was mediated through change in self-efficacy and family support. Five behavioral counseling sessions maintained self-efficacy and increased family and friend support although the intervention did not directly affect activity. The intervention can be replicated within various community settings.