Cardiac arrest in patients with heart failure may be the result of remediable factors such as pulmonary edema, drug toxicity, or electrolyte abnormalities, or it may be due to primary arrhythmias. The relation of prior aborted cardiac arrest or sustained ventricular tachycardia to subsequent prognosis was assessed in 458 consecutive patients referred for management of advanced heart failure (left ventricular ejection fraction 0.2 +/- 0.07). All patients received tailored vasodilator and diuretic therapy and were then followed as outpatients. Patients were divided into four groups: 388 patients (85%) with no prior cardiac arrest or sustained ventricular tachycardia, 31 patients (7%) with a primary arrhythmia cardiac arrest, 22 patients (5%) with a secondary cardiac arrest, and 17 patients (4%) with sustained ventricular tachycardia without cardiac arrest. Patients with cardiac arrest resulting from a primary arrhythmia were usually treated with antiarrhythmic drugs (25 patients), and five patients received an implantable defibrillator. After hospital discharge actuarial 1-year sudden death risk (17%) and total mortality (24%) rates for the group with primary arrhythmia were similar to corresponding values in patients with no history of cardiac arrest or sustained ventricular tachycardia (17% and 30%, respectively). In patients with a secondary cardiac arrest as a result of exacerbation of heart failure (11 patients), torsade de pointes (10 patients), or hypokalemia (one patient), therapy focused on removal of aggravating factors. Actuarial 1-year sudden death (39%) and total mortality (54%) rates for the group with secondary arrest were higher than for patients without a history of cardiac arrest (p = 0.003 and 0.005, respectively).(ABSTRACT TRUNCATED AT 250 WORDS)