Seventy-four patients (16 women, 58 men, age 58 +/- 11 years, mean +/- standard deviation) who received an implantable cardioverter defibrillator (ICD) after experiencing a single episode of ventricular tachycardia or ventricular fibrillation were followed to determine if antiarrhythmic drug therapy affects the incidence of ICD discharges. Thirty-three patients (group A) were treated with an antiarrhythmic drug that was either untested or previously demonstrated during electropharmacological testing to be ineffective in suppressing the induction of ventricular tachycardia. Forty-one patients (group B) were not treated with an antiarrhythmic drug. There were no significant differences between the two groups in regards to age, sex, incidence of coronary artery disease, left ventricular function or the type of ICD pulse generator used. During a mean follow-up of 14 months for the entire cohort, 15 patients (46%) in group A and 18 patients (44%) in group B experienced at least one ICD shock. The time to the first appropriate shock (5 +/- 5 months in both groups) and the frequency of ICD shocks (0.3 +/- 0.2/month in group A vs 0.4 +/- 0.5/month in group B) were similar in both groups. The incidence of syncope at the time of ICD discharge was higher in group A than group B patients (31% vs 5%, P less than 0.05). In conclusion, antiarrhythmic drugs that are untested or have failed electropharmacological testing do not appear to reduce the probability of ICD discharge over a short-term (mean 14 months) follow-up in patients who have had only one clinical episode of VT/VF and may increase the risk of syncope during ICD discharge.(ABSTRACT TRUNCATED AT 250 WORDS)