To determine the effect of a prior internal mammary artery graft on coronary artery bypass reoperation, experience with 189 consecutive patients who underwent such surgery was reviewed. Some 147 patients (group I) received only saphenous vein grafts at the primary coronary bypass surgery (CABG) and 42 (group II) received at least one IMA graft at the primary CABG. There were no differences in preoperative patient characteristics or operative data between the groups. Significantly more redo CABG 0-5 years after the initial operation was seen in group II compared with that in group I, indicating inadequate first operation or technical difficulties. In group II a larger proportion of the patients had patent grafts at redo (52.4% versus 34.7%). There were no entry injuries to the grafts or the heart in either group. No operative mortality was encountered in group II, while seven patients in group I died (P < 0.05). Group II had more pneumonia (P < 0.01) and re-exploration for bleeding (P < 0.001) than group I. However, the overall postoperative morbidity in group II patients was less than in group I, though not statistically significant. When comparing patients with an occluded internal mammary artery graft at redo (group A) with those who had a patent internal mammary artery graft (group B) there were no statistically significant differences in patient characteristics and preoperative patient profile, even though group B patients showed a trend towards a better preoperative cardiac profile. A mean of 2.4 grafts/patient were performed in group B compared with 4.0 in group A (P < 0.01). Other operative parameters did not differ between the groups.(ABSTRACT TRUNCATED AT 250 WORDS)