Patient recruitment for a small randomized study and several unrandomized studies conducted by the Thrombolysis in Myocardial Infarction investigators during 1984, 1985, and early 1986 will be reviewed. These studies, with patient numbers ranging from 48 to 316, were conducted to obtain information for planning a large-scale trial of thrombolytic therapy. Recruitment success of the 13 clinics was measured by comparing actual recruitment with the expectation of one patient per week and by comparing actual recruitment with the clinic directors' projections. Both methods indicated that four clinics had recruitment above expectation and five clinics had recruitment below expectation. Clinic staff members were requested to maintain logs of patients entering the coronary care unit and to complete screening forms on patients who met certain criteria but were ineligible. Considerable variation was noted in the number of patients screened and the number eligible.