In a series of 100 consecutive patients the internal mammary arteries have been used whenever possible for revascularization of the coronary arteries. At least one internal mammary artery was used in 89% of cases; both internal mammary arteries were used in 30% and in 17% sequential internal mammary-coronary artery anastomoses were performed. The results in this group of patients were compared to another group of 250 consecutive patients in whom the coronary arteries were revascularized by classical saphenous vein grafts. In-hospital mortality was 3% for the internal mammary artery group and 2.5% for the saphenous vein group (NS). The follow-up was complete for 89% of the patients after a mean duration of 6 months. At control the mean NYHA functional class was assessed as 1.0 +/- 0.2 for the internal mammary artery group and 1.2 +/- 0.2 for the saphenous vein group (p less than 0.01). Stress test was electrically positive in 7.1% of the internal mammary artery group and in 15.3% of the saphenous vein group (p less than 0.05). Although the method is technically more demanding, the results after internal mammary-coronary artery anastomoses appear to be superior.