The quantitative benefit of ICD therapy in patients with malignant ventricular tachyarrhythmia with different degrees of left ventricular dysfunction is unclear. We evaluated patterns of ICD use and survival in 112 patients with moderate to severe left ventricular dysfunction. Group 1 included 57 patients with moderate left ventricular dysfunction (defined as left ventricular ejection fraction greater than 30%) and group 2 comprised 55 patients with severe left ventricular dysfunction (defined as ejection fraction equal to or less than 30%). The follow-up period ranged from 1 to 78 months. Age, incidence of coronary artery disease, and presenting arrhythmia in the two groups were similar. The mean left ventricular ejection fraction in group 1 was 44.6 +/- 8.2% and in group 2 was 21.6 +/- 6% (p < 0.0001). At 3 years of follow-up 65% of the patients in group 1 and 71% in group 2 (p = NS) had ICD activation for presumed ventricular tachycardia. Survival was calculated by means of actuarial analysis. Arrhythmia or sudden death mortality at 4 years of follow-up was 5% in group 1 and 9% in group 2 (NS). Cardiac mortality was higher in patients with severe left ventricular dysfunction reaching levels of statistical significance at 2 years of follow-up. At 2 years of follow-up it was 12% in group 1 and 40% in group 2 (p = 0.05), and at 4 years of follow-up it was 15% in group 1 and 43% in group 2 (p < 0.01). In both groups there was no difference in cardiac mortality in patients who did and did not have appropriate ICD shocks.(ABSTRACT TRUNCATED AT 250 WORDS)