The objective of the present study was to analyze the autonomic response of trained swimmers to traditional and reverse training periodization models. Seventeen swimmers were divided in two groups, performing a traditional periodization (TPG) or a reverse periodization (RPG) during a period of 10 weeks. Heart rate variability and 50 m swimming performance were analyzed before and after the training programs. After training, the TPG decreased the values of the high frequency band (HF), the number of differences between adjacent normal R-R intervals longer than 50 ms (NN50) and the percentage of differences between adjacent normal R-R intervals more than 50 ms (pNN50), and the RPG increased the values of HF and square root of the mean of the sum of the squared differences between adjacent normal R-R intervals (RMSSD). None of the groups improved significantly their performance in the 50-m test. The autonomic response of swimmers was different depending on the periodization performed, with the reverse periodization model leading to higher autonomic adaption. Complementary, the data suggests that autonomic adaptations were not critical for the 50-m swimming performance.