The aim of this study was to investigate the training-induced effect on cardiac parasympathetic nervous system (PNS) activity, assessed by resting heart rate variability (HRV) and post-exercise heart rate recovery (HRR), in sedentary healthy people. Electronic searches were carried out in PubMed, Embase, and Web of Science. Random-effects models of between-group standardised mean difference (SMD) were estimated. Heterogeneity analyses were performed by means of the chi-square test and I(2) index. Subgroup analyses and meta-regressions were performed to investigate the influence of potential moderator variables on the training-induced effect. The results showed a small increase in RMSSD (SMD(+) = 0.57 [95% confidence interval (CI) = 0.23, 0.91]) and high frequency (HF) (SMD(+) = 0.21 [95% CI = 0.01, 0.42]) in favour of the intervention group. Heterogeneity tests reached statistical significance for RMSSD and HF (p </= 0.001), and the inconsistency was moderate (I(2) = 68% and 60%, respectively). We found higher training-induced effects on HF in studies that performed a shorter intervention or lower number of exercise sessions (p </= 0.001). Data were insufficient to investigate the effect of exercise training on HRR. Exercise training increases cardiac PNS modulation in sedentary people, while its effect on PNS tone requires future study.