Mammography remains substantially under-used in low-income minority populations despite its well-established efficacy as a means of breast cancer control. The Metropolitan Detroit Avoidable Mortality Project is a 2-year controlled clinical trial of coordinated interventions which seek to improve the use of early breast cancer detection services at five clinical sites providing primary health care services to inner-city women. Baseline assessment for two of the five participating clinic populations demonstrated that only one-quarter of women who visited these clinics were referred for mammography in 1988, and only half of those who were referred were able to complete the procedure. Patient characteristics including age, marital status, ethnicity, and insurance status were not associated with use of mammography during the baseline period. Each of the project's intervention components is a cue to action: a physician prompt for mammography referral within the medical record of procedure-due women, a reminder postcard for scheduled appointments, and a telephone call to encourage rescheduling of missed appointments. The interventions are initiated by a computerized information management system in the existing network of health care services. The patient's out-of-pocket mammography expense has been eliminated in three of the five sites. Although their efficacy as individual interventions has been well established, a controlled trial of computer prompts to physicians, reduced expense for patients, and patient appointment reminders as an integrated system in inner-city medical care settings has not been previously described. We have implemented the prompting, facilitated rescheduling procedures, and eliminated patient expense for mammography at three of five eventual clinical sites. This report provides an overview of the study's design, data management system, and methodology for evaluation.