Classic risk factors for mortality due to chronic heart failure (CHF), such as low left ventricular ejection fraction, NYHA functional stage, and increased heart rate perform well in the prediction of death from pump failure. The prediction of sudden cardiac death (SCD) remains somewhat problematic. Numerous studies have analyzed the potential contribution heart rate variability (HRV) can make to risk assessment in CHF. The aim of this review was to summarize the literature and identify the role HRV might play in identifying mode of death, as well as overall mortality risk. In studies where all-cause mortality or cardiac events were the clinical end point(s), global and slow oscillatory measures of HRV were the strongest risk predictors. In the fewer studies that used SCD as an end point, the strongest risk factors were HRV measures of short-term oscillations and sympathovagal interaction. We concluded from these findings that different HRV measurements predict different modes of death in CHF. Additionally, further studies using short-term analysis of HRV and non-linear analyses are warranted. Furthermore, studies with multiple end points, which clearly delineate pump failure from SCD, may be useful to identify more clearly the role HRV measures can play in the prediction of SCD.