Cancer control investigations in the United States rarely use population-based registries as a resource for recruitment. A placebo-controlled, double-blind, randomized toxicity trial of tamoxifen was conducted among postmenopausal women with node negative breast cancer. To achieve the accrual goal of 140 subjects in this single institution study, the Wisconsin Cancer Reporting System (WCRS), a population-based cancer registry, was used. Registry information from the last 9 years was used to identify 3,585 women who met the study criteria with respect to age, stage, and previous therapy. The vital status of identified women was confirmed using Wisconsin state death records. For identified cases, rosters were prepared and sent to the physician with a cover letter and study description. The physicians were asked to update the list and, if appropriate, to sign letters to potentially eligible and interested women. Thirty-eight percent of women receiving a letter and study information from their physicians contacted the study office about participation. Eighteen months from its initiation, 140 women were entered on study. This successful use of a population-based cancer registry illustrates an efficient recruitment method which could be modified for other cancer control/chemoprevention trials.