Formal evaluation of recruitment strategies for clinical trials conducted in the community setting, rather than in hospitals, is limited despite recruitment rates and costs having significant implications for the planning of large trials. As part of the development of a large-scale primary prevention trial of low-dose aspirin in 15,000 elderly persons (aged 70 years and above) to be undertaken in a community setting, a pilot study involving 400 participants was carried out to evaluate different recruitment strategies. Three techniques were compared: recruiting from the electoral roll (by individual invitation using direct mail), local community approach (through retirement villages, local media, elderly community groups), and using general/family practice (identifying persons over 70 within a practice and direct mail invitation to participate). Screening of potential participants was carried out by a trained trial nurse with limited involvement only of medical practitioners. The target recruitment figure was achieved in 4 months using the equivalent of 1.5 full-time nurses. General practice was the most effective method utilized, with a yield of one participant recruited per six approached at an estimated cost of A$48.36 per recruited participant. This compares with yield rates of 1 per 29 (for the local community approach) and 1 per 17 for the electoral roll, and costs per participant of A$42.54 and A$59.37 respectively. Other less tangible benefits of general practice include the greater cooperation of primary care practitioners in patient follow-up and end-point ascertainment.