<p>Exercise training as a potential intervention to improve male reproductive health, its efficacy appears to be contingent on exercise intensity. However, the optimal training intensity remains poorly defined. Hence, this study aimed to identify the optimal exercise intensity for improving sperm parameters through network meta-analysis of rodent studies. The comprehensive search across eight databases ultimately identified 35 eligible studies, coving four different exercises: low-intensity continuous training (LICT), moderate-intensity continuous training (MICT), high-intensity continuous training (HICT), and high-intensity interval training (HIIT). The network meta-analysis indicated that MICT showed superior efficacy versus HICT in terms of sperm concentration (MD: 15.28, 95% CI: 2.52–28.04). The surface under the cumulative ranking curve (SUCRA) value, an indicator to quantify the relative ranking probability of interventions, indicated the highest probability of MICT being optimal training for increasing sperm concentration in both healthy (SUCRA 77.6%) and unhealthy groups (SUCRA 80.1%). Additionally, MICT (MD: 12.23, 95% CI: 7.97–16.48) and HIIT (MD: 10.94, 95% CI: 1.10, 20.78) were superior to sedentary control in improving sperm motility. However, HICT was associated with a significant reduction in normal sperm morphology compared to sedentary control (MD: -10.05, 95% CI: -19.53- -0.58) and MICT (MD: -17.25, 95% CI: -32.17- -2.33). These preclinical results position MICT as the preferred exercise modality for sperm quality improvement, with HIIT as an alternative. However, environmental and lifestyle confounders unique to humans require careful attention when extrapolating these findings to human populations.</p>