To study the effect of various perioperative and operative variables, we analysed the results of 66 consecutive patients undergoing mitral valve replacement (MVR) and coronary artery bypass grafting (CABG). The mean age was 61.2 years (34 males and 32 females), the mean follow-up 54.71 +/- 7.8 months. The hospital mortality rate was 7.6% (5/66). New York Heart Association (NYHA) functional class (P < 0.01), left ventricular global wall motion score (increased scores indicating impaired function, P = 0.005) and cross-clamp time (P < 0.05) were associated with hospital mortality. There was no significant relationship of age (certainly up to the age of 70), cause of mitral valve disease, severity of mitral regurgitation, number of grafts, presence of angina, or previous myocardial infarction with hospital mortality. There were eight late deaths, survival at 1, 3 and 5 years was 92.4%, 83.2% and 80.2%, respectively. Although there was a trend for pulmonary vascular resistance (P = 0.15), NYHA class (P = 0.18) and aortic cross-clamp time (P = 0.09) to be associated with late survival, the only factor significantly related to late survival was global wall motion score (P = 0.001), i.e. those with scores of more than 10. Severity of mitral regurgitation and cause of mitral valve disease have been reported as being related to late survival in patients undergoing combined CABG and MVR, but we have found no such relationship. Our results indicate that both hospital and late mortality after this operation are strongly correlated with left ventricular function.