Objective: To describe the pilot and early implementation phase of a system for assembling and recruiting cohorts of patients taking selected prescription medications and prospectively monitoring them for new health events. Design: Prospective observational study, based on telephone interviews, of 1475 patients filling prescriptions for a nonsteroidal antiinflammatory drug (NSAID). Patients were interviewed by telephone using trained interviewers at a central site. Hospitalizations and deaths were followed up and reviewed by an independent physician. Setting: Community setting in a region of Hamilton, Ontario, Canada. Participants: All consenting patients filling new or repeat prescriptions for NSAIDs at participating pharmacies. Main Outcome Measures: The authors report on the development and assessment of systems for: (1) ongoing recruitment of patients through community pharmacies; (2) data transfer from pharmacies to the coordinating center; (3) surveying patients; (4) classifying, coding, and evaluating new health events; and (5) following up on new serious adverse events. Results: Fifty-one percent of patients approached were recruited, and 83% of these provided completed interviews. For patients picking up their own medications, pharmacy workload varied from 4 to 10 minutes per patient approached. Nineteen percent of patients reported having a new health problem or unusual symptom at the initial telephone interview. Reported health-related events were similar to those described in other studies of NSAIDs. Conclusions: Most aspects of the monitoring system performed well. One limitation was the low recruitment rate for patients who did not directly drop off or pick up their own prescriptions. Even so, this method of patient accrual may complement alternative monitoring programs.