OBJECTIVE: To analyse the relation between perioperative myocardial injury (PMI) and the risk of subsequent heart failure after coronary artery bypass grafting (CABG). DESIGN AND SETTING: Clinical data were documented prospectively in all patients and stored in a computer. All hospital readmissions were identified and the registered primary diagnoses were analysed. Survival information on all patients was obtained by use of combined registers. The study was carried out at the cardiac surgical referral centre of University Hospital, Uppsala, Sweden. PATIENTS: 7493 patients discharged alive after primary CABG between 1987 and 1996 were followed up until the first hospital readmission for heart failure, death, or 31 December 1996 was reached. MAIN OUTCOME MEASURES: Hospital readmission for heart failure or late mortality. RESULTS: Of the patients studied 576 (7.7%) were readmitted for heart failure. Actuarial freedom from readmission for heart failure after four years was 93%, and after seven years, 89%. Of the 576 patients, 114 (20%) had had PMI, which increased the risk of heart failure independently (hazard ratio (HR) 2.3, 95% confidence interval (CI) 1.8 to 2.8). Increased age, female sex, diabetes, previous myocardial infarction, dyspnoea, preoperative atrial fibrillation, left ventricular dysfunction, and triple vessel disease were independent risk factors for heart failure. The use of an internal mammary artery decreased the risk. PMI implied increased mortality (HR 1.4, 95% CI 1.1 to 1.8). Late mortality was greatly increased in patients readmitted for heart failure. CONCLUSION: PMI increased the risk of heart failure and late death after CABG, and heart failure had a notable adverse effect on late survival.