Sixty-two patients underwent 65 heart-lung transplant operations before April 1989. Thirty of these patients (operated on March 1981 to February 1986) were given cyclosporine and prednisone for immunosuppression. These 30 patients (group 1) are compared with the 32 most recent patients (group 2) treated with cyclosporine, prednisone, and azathioprine. Patient characteristics (group 1 versus group 2) were similar including age (mean 32 versus 26 years) and indication for operation (51% versus 38%, primary pulmonary hypertension; 46% versus 45%, Eisenmenger complex; 3% versus 17%, other). The perioperative (in-hospital) mortality rate was 35% in group 1 versus 16% in group 2. Obliterative bronchiolitis developed in 12 (63%) of 19 hospital survivors in group 1 from 2 months to 49 months postoperatively (mean 12 months). This complication has developed in five (20%) of the 25 hospital survivors in group 2 from 3 to 13 months postoperatively (mean 8 months). Overall, 24 of the 30 patients in group 1 died (infection in eight, obliterative bronchiolitis in five, graft coronary artery disease in two, other causes in nine). Eight of the 32 patients in group 2 died (infection in six, other cases in two). Survival rates for group 1 patients were 60% at 1 year, 50% at 2 years, 43% at 3 years, and 25% at 5 years. Survival rates for group 2 patients were 73% at 1 year, 73% at 2 years, and 65% at 3 years (p less than 0.05). When group 2 patients were compared with 134 patients undergoing heart transplantation during the same time period, there was no difference in survival. Routine bronchoscopic surveillance in the more recent group of patients has led to the earlier detection of rejection and infection. In combination with decreased perioperative mortality and augmented immunosuppression, this had led to improved survival in our more recent heart-lung transplant recipients. In addition, the earlier recognition of infection and rejection has led to a decrease in severity of obliterative bronchiolitis.