During a 9-year period 80 patients with drug-resistant sustained ventricular tachycardia underwent direct surgical ablation of arrhythmogenic myocardium. Sixty-nine were male (86%) and 11 female (14%), with 1.9 +/- 1.1 (standard deviation) ventricular tachycardia morphologies per patient. The mean number of drugs failed was 3.7 +/- 1.6 per patient. The preoperative left ventricular ejection fraction was 36.4% +/- 14.4%. Complete preoperative endocardial mapping data (greater than 4 endocardial sites in each ventricular tachycardia) were available for 60 of the 80 patients (75%) and intraoperative endocardial data in the clinical ventricular tachycardia was obtained in 37 (46.3%) of the patients. In 17 patients mapped intraoperatively by computer-assisted techniques, complete epicardial and endocardial data in the clinical ventricular tachycardia were obtained in 14 patients (82.4%). Overall, 73 of 80 (91.3%) had some mapping data available. Hospital mortality occurred in 10 patients (12.5%) at a mean interval of 13.5 days, range 0 to 62 days. Postoperatively the clinical ventricular tachycardia has not recurred in 65 of 70 surviving patients (92.9%). Nonclinical ventricular tachycardia occurred in another four patients. All nine patients with postoperative ventricular tachycardia responded to drugs. The major factors predictive of hospital mortality were prolonged cardiopulmonary bypass (greater than 150 minutes), preoperative ejection fraction less than 31%, and incomplete preoperative mapping. Hospital mortality in patients with an ejection fraction below 31% was significantly associated with a history of amiodarone usage. At 3 years of follow-up, freedom from sudden cardiac death was 95.7%, and 86.7% of patients were free of ventricular tachycardia on no antiarrhythmic drugs. These results suggest that direct ventricular tachycardia operations are an effective form of therapy for patients with sustained monomorphic ventricular tachycardia.