OBJECTIVES: To systematically perform a meta-analysis of the scientific literature to determine whether the outcomes of endurance training based on heart rate variability (HRV) are more favorable than those of predefined training. DESIGN: Systematic review and meta-analysis. METHODS: PubMed and Web of Science were searched systematically in March of 2020 using keywords related to endurance, the ANS, and training. To compare the outcomes of HRV-guided and predefined training, Hedges\' g effect size and associated 95% confidence intervals were calculated. RESULTS: A total of 8 studies (198 participants) were identified comprising 9 interventions involving a variety of approaches. Compared to predefined training, most HRV-guided interventions included fewer moderate- and/or high-intensity training sessions. Fixed effects meta-analysis revealed a significant medium-sized positive effect of HRV-guided training on submaximal physiological parameters (g鈥?鈥?.296, 95% CI 0.031 to 0.562, p鈥?鈥?.028), but its effects on performance (g鈥?鈥?.079, 95% CI -0.050 to 0.393, p鈥?鈥?.597) and V虈O2peak (g鈥?鈥?.171, 95% CI -0.213 to 0.371, p鈥?鈥?.130) were small and not statistically significant. Moreover, with regards to performance, HRV-guided training was associated with fewer non-responders and more positive responders. CONCLUSIONS: In comparison to predefined training, HRV-guided endurance training had a medium-sized effect on submaximal physiological parameters, but only a small and non-significant influence on performance and V虈O(2peak). There were fewer non-responders regarding performance with HRV-based training.