PURPOSE: The purpose of this study is to evaluate the effects of an 8-week exercise-based cardiac rehabilitation program on traditional and nonlinear heart rate variability (HRV) indexes, assessing the potential confounding influences of habitual physical activity (PA) and dietary intake. METHODS: In this parallel-group trial, 96 patients (56 +/- 10 years old) were randomized to the exercise group (EG) or to the control group (CG) 4 weeks after an acute myocardial infarction. Exercise-based cardiac rehabilitation program consisted of aerobic exercise at 70% to 85% of maximal heart rate for 3 sessions per week plus usual care. The CG received only usual care. The baseline and final assessments comprised resting short-term HRV (primary outcome) by a Polar R-R recorder under controlled breathing (12 breaths per minute), habitual PA by accelerometers, and dietary intake by a 4-day food diary. RESULTS: Two patients in each group dropped out and were not included in the intention-to-treat analysis. In the remaining 92 patients (EG = 47 and CG = 45), at baseline, only a difference in the proportion of nitrate medication use was significant between groups. After 8 weeks, no significant changes were found between groups on traditional and nonlinear HRV indexes (eg, ln HF, EG from 5.7 +/- 1.5 to 5.9 +/- 1.3 and CG from 5.5 +/- 1.6 to 5.5 +/- 1.5), habitual PA, and dietary intake. CONCLUSION: Eight weeks of exercise-based cardiac rehabilitation program is an insufficient stimulus to improve cardiac autonomic function in post-myocardial infarction patients under optimal medication and with high levels of traditional and nonlinear HRV indexes at baseline.