To assess the prevalence and significance of left ventricular dilatation in patients with severe left ventricular dysfunction secondary to coronary artery disease (or coronary artery cardiomyopathy), we studied 70 patients with an ejection fraction of 35 percent or less and one-vessel coronary artery disease (n = 14) or with multivessel coronary artery disease (n = 56). None had a recent myocardial infarction or valvular heart disease. Patients who underwent myocardial revascularization during follow-up were excluded. The left ventricular end-diastolic volume (measured by contrast ventriculography) was less than 110 ml/sq m in 14 patients (20 percent) (group 1), and was 110 ml/sq m or more in 56 patients (80 percent) (group 2). There were no differences between the two groups in age, sex, diabetes mellitus, hypertension, extent of coronary artery disease, or left ventricular asynergy. Patients in group 1 had lower pulmonary arterial wedge pressure (13 ± 6 vs 22 ± 10 mm Hg; p = 0.0008), lower left ventricular end-diastolic pressure (21 ± 6 vs 27 ± 9 mm Hg; p = 0.007), and higher left ventricular ejection fraction (31 ± 2 vs 25 ± 7 percent; p = 0.001) than patients in group 2. At a mean follow-up of 27 months, 24 patients had died of cardiac causes, all of whom were in group 2. Survival was significantly better in group 1 than in group 2 (Mantel-Cox, p = 0.009). Survival analysis (Cox model) of 20 clinical, hemodynamic, and angiographic variables showed that ejection fraction (χ2 = 13.6; p < 0.001) and end-diastolic volume χ2 = 4.7; p = 0.03) were the most significant predictors of death. Thus, minimally dilated coronary artery cardiomyopathy is a distinct entity with favorable hemodynamics. Prognostically, the end-diastolic volume adds significant predictive information to the ejection fraction among conservatively treated patients.