There is a delayed (lag 1 to 2 days) correlation between acute PM 2.5 (particulate matter < 2.5 mum in aerodynamic diameter) exposure and cardiovascular events, but the underlying mechanism remained unclear. We aimed to investigate the delayed impact of acute PM 2.5 exposures on cardiac autonomics through linear and nonlinear heart rate variability (HRV) analyses. Among 6912 patients who had received 24-h Holter ECG between October 1, 2015, to October 31, 2016, 56 patients (31 males, 70.3 +/- 12.7 years old) were enrolled. We classified the patients as high (> 35.4 mug/m(3)) or low (< 35.4 mug/m(3)) PM 2.5 groups according to their PM 2.5 exposures on the day of Holter recordings (day 0) lag 1 and lag 2 days. Linear and nonlinear HRV parametersleft broken bracketDetrended fluctuation analysis (DFA) slopes 1 and 2right broken bracketwere compared. Baseline characteristics were similar between groups. Linear and nonlinear HRV parameters were similar between high- and low-exposure groups on day 0 and lag 1 day, respectively. However, DFA slope 1 was significantly lower in the high-exposure group on lag 2 days (0.784 +/- 0.201 vs. 0.964 +/- 0.274, p = 0.021). DFA slope 1 of the high-exposure group was significantly lower on daytime periods (9 am to 9 pm, 8 am to 4 pm and 4 pm to 12 pm) but not on nighttime periods. High lag 2 days PM 2.5 exposure is associated with low DFA slope 1 and the relationship is diurnal. This suggests that air pollution might have a delayed impact on the cardiovascular autonomic system.