Although many evidence-based interventions are well-established, our understanding of how to effectively implement and sustain those interventions in real-world settings is less well understood. We investigated predictors of implementation and reach in a randomized controlled trial of the NORTH STAR prevention system. One-third of U.S. Air Force (AF) bases worldwide were randomly assigned to NORTH STAR (n = 12) or an assessment-and-feedback-only condition (n = 12). Process data regarding implementation factors were collected from Community Action Team (CAT) members and observations of CAT processes. Results from a series of regression analyses indicated that change in leadership and community support, action planning processes, and perceived approach effectiveness from pre-action planning to follow-up predicted community action plan (CAP) implementation and that changes in barriers to implementation predicted CAP reach. Pre-action planning reports of CAT member self-efficacy and perceived approach effectiveness also predicted CAP implementation at 1-year follow-up. Future directions and practice recommendations are provided.