The long-term follow-up of 52 pts (36 M, 16 F, mean age: 62 years) with sustained ventricular tachyarrhythmias (SVT) was analyzed to assess the efficacy and feasibility of empiric amiodarone treatment. Forty-five pts had organic heart disease (mean EF: 38.3%) and 7 pts no overt heart disease. Twenty pts suffered from syncope or cardiac arrest secondary to sustained ventricular tachyarrhythmias (mean: 2.35 episodes) and 32 did not. All pts were given amiodarone empirically (mean dose: 390 mg) and followed-up for a mean period of 29.5 months (range 1-137). Two pts (3.8%) died of non cardiac causes, 5 (9.6%) of non sudden cardiac death and 7 (13.4%) of sudden death. Fifteen pts (28.8%) experienced non fatal arrhythmic recurrences. Four out of 7 pts who died suddenly experienced non fatal arrhythmic recurrence before death. The actuarial incidence of cardiac death was 10.8, 22.7, 31.5, 31.5% at 1, 2, 3 and 5 years; the actuarial incidence of sudden death was 8.9, 12, 22.1, 22.1% at 1, 2, 3 and 5 years; the actuarial incidence of non fatal arrhythmic recurrences was 17.4, 26.3, 26.3, 26.3, 44.7% at 1, 2, 3, 4 and 5 years. Univariate analysis identified recent myocardial infarction, NYHA functional class, detection of frequent and/or repetitive premature ventricular contractions on Holter monitoring and non fatal arrhythmic recurrences as predictors of cardiac death (p less than 0.05), while only non fatal arrhythmic recurrences were associated with sudden death (p less than 0.05). Twenty-two pts (42.3%) developed side effects. Nine (17.3%) discontinued amiodarone: 6 pts (11.5%) because of side effects and 3 inadvertently.(ABSTRACT TRUNCATED AT 250 WORDS)