BACKGROUND: In contrast to patients with moderate to severe chronic heart failure (CHF), data regarding long-term outcome in patients with mild CHF are scarce. We examined the place of Holter monitoring to study the prognostic value of ventricular arrhythmias and heart rate variability (HRV) in patients with mild to moderate CHF during long-term follow-up. METHODS: We studied 90 patients with mild to moderate CHF and NYHA class II who had been enrolled in the Dutch Ibopamine Multicenter Trial. At baseline their mean age was 60.5 +/- 8.0 years, left ventricular ejection fraction (LVEF) was 0.29 +/- 0.09, and 85% were males. At the start of the study, patients were only using diuretics, while digoxin, and particularly ACE inhibitors and beta-blockers were initiated later. Univariate and multivariate proportional hazard analyses were performed. RESULTS: At baseline 80% of patients were in NYHA class II, and 20% were in class III; their mean age was 60 years, mean LVEF was 0.29, and 85% were men. During a follow-up of 13 years, 47 patients (53%) died. Cardiovascular (CV) death occurred in 39 patients, of which 28 were sudden cardiac death (SCD). For both CV death and SCD, LVEF <30% and ventricular premature beats/h (>20) were independent risk markers. Of the HRV parameters, total power (>2,500 ms(2)) was an important risk marker for CV death, but not for SCD. CONCLUSION: The present 13-year follow-up study in 90 patients with mild to moderate CHF showed that ventricular premature beats and HRV may have important value in predicting outcome.