Community coalitions have the potential to enhance a community's capacity to engage in effective problem solving for a range of community concerns. Although numerous studies have documented correlations between member engagement and coalition processes and structural characteristics, fewer have examined associations between coalition factors and community capacity outcomes. The current study uses data from an evaluation of the California Healthy Cities and Communities program to examine pathways between coalition factors (i.e. membership, processes), member engagement (i.e. participation, satisfaction) and community capacity as hypothesized by the Community Coalition Action Theory (CCAT). Surveys were completed by 231 members of 19 healthy cities and communities coalitions. Multilevel mediation analyses were used to examine possible mediating effects of member engagement on three community capacity indicators: new skills, sense of community and social capital. Results generally supported CCAT. Member engagement mediated the effects of leadership and staffing on community capacity outcomes. Results also showed that member engagement mediated several relationships between process variables (i.e. task focus, cohesion) and community capacity, but several unmediated direct effects were also observed. This suggests that although member engagement does explain some relationships, it alone is not sufficient to explain how coalition processes influence indicators of community capacity.