Purpose: To investigate the effects of caffeine ingestion on heart rate variability (HRV) and repeated high-intensity sprints performance in recreational athlete. Methods: Fourteen university males with a background in team-sport training and competition participated in this randomized grouping, double-blinded and crossover experimental design. Participants performed the exercise test on a cycling ergometer after ingestion of either caffeine (CAF, 6 mg•kg^(-1)) or placebo (PLA). The exercise was consisted 24×4-s sprints with 90-s active recovery. The HRV was evaluated in supine position for 60 minutes after the administration of supplementation. Results: Standard deviation of normal-normal RR interval in the CAF trial (143.83±42.45 ms) was significantly higher than in the PLA trial (108.03±31.37 ms, p＜.05). No significant differences were observed in low frequency (LF), high frequency (HF), LF/HF ratio, and coefficients of variation in HF between CAF and PLA trials. However, the coefficient of variation in LF was significantly higher in the CAF trial (5.05±1.91 %) compared with PLA trial (3.87±1.42 %, p＜.05). CAF trial resulted in 2.82% increases in peak power (p＜.001) and 3.32 % increases in mean power (p＜.001) compared with PLA trial. Conclusion: Ingestion of caffeine prior to vigorous exercise slightly increases autonomic nervous activity, and improves sprint performance during repeated high-intensity test.