In a randomized, controlled trial (N = 8,298) the authors compared three ways of encouraging patients in a large family practice to obtain a blood pressure check. Working from computerized medical records at a teaching family-medicine center, the study included patients aged 18 years and over who had not had a blood pressure measurement during the previous year. In a normal-care control group, 21.1% of those due for a blood pressure reading obtained one during the trial year. In one intervention group, a computer-generated message reminded the doctor to check the blood pressure of patients who happened to visit the center for an appointment; 30.7% of patients allocated to this group were checked. In the second group, the nurse contacted patients by telephone and encouraged 24.1% of those who were due for a check to obtain one. Sending a letter to patients in the third group yielded a 35.7% compliance rate. Considering the costs of the three interventions, the physician reminder was the most cost effective, followed by the letter reminder. Although statistically significant, the impact of the reminders was modest. A better approach might involve a combination of routine reminders to the physician, followed by letters to noncompliant patients.