This paper studies the factors associated with perioperative myocardial infarction after coronary bypass surgery and assesses the medium-term prognosis of these patients. Four hundred and seventy patients underwent coronary bypass surgery between January 1983 and December 1986. The appearance and persistence of pathological Q waves, absent on the preoperative ECG, was the unique criterion of perioperative infarction. This complication was observed in 36 patients (7.65%). A comparison of these patients with a random group of 144 of teh 434 patients without perioperative infarction showed that they had a higher incidence of crescendo angina (55% vs 21%; p less than 0.001), ST-T wave changes on the resting ECG (78% vs 46%; p less than 0.001) and poor distal left anterior descending network (33% vs 13%; p less than 0.001): in addition, the group with infarction had a lower left ventricular ejection fraction (0.58 vs 0.64, p less than 0.01), incomplete myocardial revascularisation procedures (58% vs 32%; p less than 0.01), longer cardiopulmonary bypass times (86 mn vs 69 mn; p less than 0.001) and longer aortic clamping times (44.5 mn vs 37.4 mn p less than 0.05). The acute phase of the perioperative infarct was characterised by a higher incidence of major cardiac complications such as low output states (30.5% vs 2.02%; p less than 0.001). The hospital mortality was higher in the infarct group (8.3% vs 2.01%) but this was not statistically significant. After an average follow-up of 44 +/- 3 months, the 5 year survival rate was 95.4 +/- 2.1 per cent in patients without infarction and 76.5 +/- 6.9 per cent in those with perioperative infarction (p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)