INTRODUCTION AND OBJECTIVES The duration of the QRS interval measured by ECG is a marker of ventricular dysfunction and indicates a poor prognosis. Its value in patients undergoing coronary revascularization surgery has not been established. METHODS The study involved 203 consecutive patients (age 64+/-9 years, 74% male) scheduled for elective coronary surgery. The maximum QRS duration measured on a preoperative 12-lead ECG was recorded. Hemodynamic instability was defined as the occurrence of cardiac death, heart failure, or a need for intravenous inotropic drugs or intra-aortic balloon counterpulsation during the postoperative period. RESULTS The occurrence of hemodynamic instability (n=94, 46%) was associated with a longer preoperative QRS duration (97.5+/-21.14 ms vs 88.5+/-16.9 ms; P=.001). The QRS duration was also longer in patients who developed heart failure (n=23; 104.3+/-22.9 ms vs. 91.1+/-18.5 ms; P=.002), needed inotropic drugs (n=77; 96.5+/-20.5 ms vs. 90.1+/-18.2 ms; P=.007) or developed postoperative atrial fibrillation (n=58; 98.2+/-23.8 ms vs. 90.4+/-17.0 ms; P=.018). Bundle branch block was associated with a greater need for intra-aortic balloon counterpulsation (29% vs 12%; P=.012) or inotropic drugs (58% vs 35%; P=.014) and a higher incidence of hemodynamic instability (69% vs 42%; P=.006). Multivariate analysis identified the following independent predictors of hemodynamic instability: QRS duration (adjusted odds ratio [OR] per 10 ms=1.49; 95% confidence interval [CI], 1.11-2; P=.007), the lack of an arterial graft (OR=3.6; 95% CI, 1.14-11.6; P=.029) and extracorporeal circulation time (OR per min=1.013; 95% CI, 1.003-1.023; P=.013). CONCLUSIONS The intraventricular conduction delay, or QRS duration, was associated with a higher risk of postoperative hemodynamic instability following coronary surgery.