We performed a prospective study in 35 untreated patients aged > or = 45 years, who had a mean sinus rate at rest of < or = 50 beats/min and/or intermittent sinoatrial block, and symptoms attributable to sinus node dysfunction. The patients were followed up for up to 4 years (mean 17 +/- 15 months). During follow-up, 20 patients (57%) had cardiovascular events that required treatment: 8 had syncope (23%); 6 had overt heart failure (17%); 4 patients had chronic atrial fibrillation (11%); and 2 patients had poorly tolerated episodes of paroxysmal tachyarrhythmias (6%). Actuarial rates of occurrence of all events were 35%, 49%, and 63%, respectively, after 1, 2, and 4 years. At univariate analysis, age > or = 65 years, end-systolic left ventricular diameter > or = 30 mm, end-diastolic left ventricular diameter > or = 52 mm, and ejection fraction < 55% were predictors of cardiovascular events. At multivariate analysis, age, end-diastolic diameter, and ejection fraction remained independent predictors of events. Actuarial rates of occurrence of syncope were 16%, 31%, and 31%, respectively, after 1, 2, and 4 years. Both univariate and multivariate predictors of syncope were history of syncope and corrected sinus node recovery > or = 800 ms. A favorable outcome was observed in the remaining 43% of patients. Thus, clinical cardiovascular events occur in most untreated sick sinus syndrome patients during long-term follow-up, even though a favorable course can be expected in 43% of patients. The outcome can be partly predicted on initial evaluation. In the patients with a favorable outcome, treatment can safely be delayed.