Background: Chronic hemodialysis (HD) patients suffer from an appallingly high cardiovascular mortality. During HD, patients are exposed to dialysate glucose, which may alter blood glucose levels and thus exert effects on the autonomic nervous system. Heart rate variability (HRV) is an established indicator of autonomic nervous system activity and a predictor of cardiovascular outcomes. This study investigated the effects of two commonly used dialysate glucose concentrations [100 mg/dl (HD100), and 200 mg/dl (HD200)] on HRV in chronic HD patients. Methods: In this prospective, randomized, controlled, single-masked, cross-over trial, subjects were randomized to receive HD100 or HD200 for a period of 3 weeks followed by a cross-over to the respective other dialysate (www.clinicaltrials.gov #NCT00618033). Blood glucose and insulin levels were measured before and after HD. Intradialytic Holter electrocardiograms were recorded and HRV time domain, frequency domain and complexity parameters analyzed. Results: Twenty-three HD patients (age 56 ± 12 years, 11 male, 14 black, 11 with diabetes) were studied. Diabetic subjects showed significantly higher serum glucose levels with HD200 as compared to HD100 (HD100: 146 ± 48 mg/dl; HD200: 192 ± 57 mg/dl; p < 0.01); this hyperglycemia was accompanied by an increase of the high-frequency band of HRV (p = 0.019), a reflection of increased parasympathetic activity. HRV did not change in nondiabetic subjects. Conclusion: In diabetic subjects, the use of HD200 increased vagal tone. Given the importance of sympathetic activation to counteract intradialytic hypotension, our findings support the use of HD100 in diabetic HD patients.