This study examined the usefulness of the electrophysiologic approach for selecting antiarrhythmic drug therapy to improve survival in patients with ventricular tachycardia (VT) and Chagas' disease. A total of 71 consecutive chagasic patients undergoing treatment and evaluation of VT were analyzed. Programmed electrical stimulation (PES) was performed in 45 patients, sustained VT was induced in 18 of these 45 (40%); nonsustained VT was induced in 17 (38%), and in 10 patients (22%) VT was not induced at all. An average of 3 drugs per patient were tested, including mexiletine, flecainide and propafenone. At least 1 effective drug preventing VT induction was identified in 13 of 18 patients with induced sustained VT, whose outcome resulted in 2 nonsudden but cardiac deaths (15%). Eight patients received no drug therapy because the induced arrhythmia was asymptomatic nonsustained VT; none of these died. The remaining 24 patients from the PES group were empirically treated with amiodarone; 7 died (4 suddenly) during follow-up (29%). A group of 26 patients (non-PES group) did not undergo electrophysiologic evaluation. In these patients, the therapy chosen was amiodarone alone or associated with mexiletine, and the incidence of death was 7 of 26 patients (27%), 3 suddenly (p less than 0.05 at 10-year survival and p = not significant at 5-year survival). It is concluded that the electrophysiologic approach improves survival in this study population, but only 29% were eligible for guided therapy.(ABSTRACT TRUNCATED AT 250 WORDS)