This study compared autonomic modulation in swimmers and non-athletes in relation to body composition. A total of 28 athletes with a mean age of 19.7 ± 2.9 years were evaluated who had at least 2 years of swimming training, trained approximately 7,000 m per day, with a frequency of 5 days per week, and who competed at national level. The control group was made up of 21 volunteers (23.0 ± 2.5 years), who did not practice regular physical activity (<2 hours per week). Body composition was estimated using dual-energy x-ray absorptiometry, and autonomic modulation was assessed by heart rate variability (HRV). The results show that there were significant differences in autonomic modulation and body composition between the groups, and that the athletes had a higher overall variability (standard deviation of all normal intervals between consecutive heart beats [SDNN]: 78.1 [72.5-93.5] × 61.1 [56.4-75.7], p = 0.022) and greater autonomic balance (LF/HF: 0.96 [0.88-1.35] × 0.71 [0.56-0.93], p = 0.023), compared with the non-athletes, respectively. In addition, a moderate and positive relation was obtained between fat-free mass and the square root of the squared differences between consecutive heartbeat intervals (RMSSD: r = 0.526, p = 0.004 × r = 0.456, p = 0.038), (SDNN: r = 0.617, p = 0.001 × r = 0.571, p = 0.007) and low frequency (LFms: r = 0.517, p = 0.005 × r = 0.600, p = 0.004) in the athletes and non-athletes, respectively, without a correlation between fat mass (FM). The conclusion is that young highly trained swimmers had lower FM, increased fat-free mass, and better HRV than young adult non-athletes and suggests that a lower quantity of FM and, especially, a greater fat-free mass (FFM) are linked to better autonomic modulation. Thus, this study could contribute to coaches and trainers establishing greater performance by better autonomic modulation and greater quantity of FFM.