This study investigated the impact of mild COVID-19 on HRV in groups stratified by time after infection and to compare to a healthy group of the same age without previous virus infection and without need of hospitalization. This is a cross-sectional study. We divided the sample into four groups: control group (CG) (n = 31), group 1 (G1): ≤6 weeks (n = 34), group 2 (G2): 2-6 months (n = 30), group 3 (G3): 7-12 months (n = 35) after infection. For HRV analysis, we used the indices of linear (time and frequency domain) and non-linear analysis. For comparisons between groups, ANOVA one way test or Kruskal-Wallis was used according to the data distribution. The effect size was calculated based on Cohen's d or η. Simple and multiple linear regressions were performed to investigate the interaction between clinical outcomes and HRV parameters. A total of 130 individuals were included. Groups G1 and G2 showed less parasympathetic modulation when compared to CG (p < 0.05), while G3 showed an increase in parasympathetic modulation when compared to G1 (p < 0.05). Moderate to large effect sizes were found according to Cohen d or η. The multiple linear regression models identified age and infection duration as significant predictors for RMSSD (adjusted R = 0.227) and SD1 (adjusted R = 0.242), while age was significant for SDNN (adjusted R = 0.213). BMI, hypertension, and dyslipidemia were non-significant in all models. For HF (n.u.), infection duration was consistently significant, with stress emerging as a predictor in Model 2 (adjusted R = 0.143). The recovery time since diagnosis and age influences recovery from HRV, suggesting a transient effect of the disease on the autonomic nervous system.