The population of patients with heart failure (HF) and mild to moderate left ventricular (LV) dysfunction is growing, and mortality remains high. There is a need for better risk stratification of patients who might benefit from primary prevention of mortality. This study aimed to evaluate the prognostic value of Holter-based parameters for predicting mortality in patients with HF with LV ejection fraction (EF) >35%. The study involved 294 patients (199 men, mean age 66 years) with HF of ischemic and nonischemic causes, New York Heart Association classes II to III, and LVEF >35%. Surface electrocardiogram and 24-hour Holter monitoring were performed at enrollment to assess traditional electrocardiographic variables, as well as heart rate variability, heart rate turbulence, and repolarization dynamics (QT/RR). Total mortality and sudden death were the primary and secondary end points. During a median 44-month follow-up, there were 43 deaths (15%). None of the traditional electrocardiographic risk parameters significantly predicted mortality. A standard deviation of all normal-to-normal RR intervals < or =86 ms, turbulence slope < or =2.5 ms/RR, and QT end/RR >0.21 at daytime were found to be independent risk predictors of mortality in multivariate analyses. The predictive score based on these 3 variables showed that patients with > or =2 abnormal risk markers were at risk of death (30% 3-year mortality rate) and sudden death (12%), similar to death rates observed in patients with LVEF < or =35%. In conclusion, increased risk of mortality and sudden death could be predicted in patients with HF with LVEF >35% by evaluating the combination of standard deviation of all normal-to-normal RR intervals, turbulence slope, and QT/RR, parameters reflecting autonomic control of the heart, baroreflex sensitivity, and repolarization dynamics.