We are conducting a community-based cluster-randomized trial in rural Alabama, testing a peer-support intervention designed to improve diabetes self-care behaviors. We describe recruitment and data collection approaches used, focusing on strategies that created community partnerships and facilitated recruitment in underserved, rural, largely minority communities. Key recruitment and data collection strategies included early community engagement; pilot testing of procedures; inclusion of community members as study team members, recruiters, and data collectors; data collection at community venues to minimize participant travel requirements; and provision of a multi-disciplinary diabetes education program to both intervention and control participants. A total of 424 participants were recruited and enrolled (400 targeted). Of the 759 referrals received, 78.9% (n= 599) successfully completed telephone screening. Of these, 78.8% (n= 472) were eligible and scheduled for a local enrollment day, and 81.4% (n= 384) attended and enrolled in the study. In addition, community members who walked in and expressed interest were screened, and 40 eligible and willing individuals were consented and enrolled.We exceeded recruitment goals in underserved, rural communities in Alabama. This success was due in large part to community partnerships that facilitated community involvement on several levels: engaging the community early in study proposal and design; hiring community members to fill various capacities as research team members, recruiters, and data collectors; conducting data collection within communities; and collecting additional contact information to maintain communication. Providing diabetes education to all participants, including intervention and control, helped ensure that everyone stood to benefit and likely enhanced overall participation. © 2012 Elsevier Inc.