Purpose of this study was to test utility of heart rate variability (HRV) in daily endurance exercise prescriptions. Twenty-six healthy, moderately fit males were randomized into predefined training group (TRA, n = 8), HRV-guided training group (HRV, n = 9), and control group (n = 9). Four-week training period consisted of running sessions lasting 40 min each at either low- or high-intensity level. TRA group trained on 6 days a week, with two sessions at low and four at high intensity. Individual training program for HRV group was based on individual changes in high-frequency R-R interval oscillations measured every morning. Increase or no change in HRV resulted in high-intensity training on that day. If there was significant decrease in HRV (below reference value [10-day mean-SD] or decreasing trend for 2 days), low-intensity training or rest was prescribed. Peak oxygen consumption (VO(2peak)) and maximal running velocity (Load(max)) were measured in maximal treadmill test before and after the training. In TRA group, Load(max) increased from 15.1 +/- 1.3 to 15.7 +/- 1.2 km h(-1) (P = 0.004), whereas VO(2peak) did not change significantly (54 +/- 4 pre and 55 +/- 3 ml kg(-1) min(-1) post, P = 0.224). In HRV group, significant increases were observed in both Load(max) (from 15.5 +/- 1.0 to 16.4 +/- 1.0 km h(-1), P < 0.001) and VO(2peak) (from 56 +/- 4 to 60 +/- 5 ml kg(-1) min(-1), P = 0.002). The change in Load(max) was significantly greater in HRV group compared to TRA group (0.5 +/- 0.4 vs. 0.9 +/- 0.2 km h(-1), P = 0.048, adjusted for baseline values). No significant differences were observed in the changes of VO(2peak) between the groups. We concluded that cardiorespiratory fitness can be improved effectively by using HRV for daily training prescription.