Padel, an increasingly popular sport, presents some associated cardiovascular risks, which can be prevented by heart rate variability (HRV) monitoring. However, there is no study worldwide that characterizes HRV during padel games. Therefore, this study aims to monitor HRV responses and characterize them using linear and non-linear metrics at three timepoints: at baseline, during a game, and during recovery. Twenty-seven amateur participants had their HRV monitored before, during, and after a 90 min padel game. We extracted time, frequency, and non-linear measures with MATLAB for baseline, rest period, and at three periods of 5 min during the game. The differences in measures were assessed through an ANOVA. The autonomic modulation was affected by the padel match in amateur players. In this regard, the root mean square of successive differences between normal heartbeats (RMSsd), Poincare plot (perpendicular standard deviation (SD1) and horizontal standard deviation (SD2)), sample entropy (SampEn), low frequencies (LF), and very low frequencies (VLF) were significantly reduced during the game, whereas alpha-2, high frequencies (HF), and the ratio between low and high frequencies (LF/HF) increased during the game. Furthermore, an abrupt change was found in the autonomic modulation between game and recovery assessments, which suggests the need to rethink the practices of cool-down protocols. The multiple timepoints analyzed during this study allow us to investigate the evolution of different HRV measures in the time, frequency, and non-linear domains, clarifying the interpretation of the variables, especially the less-investigated ones such as the non-linear measures.