This study aimed to investigate the relationship between heart rate variability (HRV), a parameter of the autonomic nervous system activity (ANSA), and postoperative delirium and postoperative events. This retrospective cohort study included elderly patients aged 65 years or older who were admitted to the intensive care unit (ICU) after cardiovascular surgery. ANSA was measured using HRV parameters for 1 h at daytime and 1 h at night-time before ICU discharge. The primary endpoint was the effect of HRV parameters and delirium on mortality and readmission rates within 1 year after discharge, and the secondary endpoint was the association between HRV parameters and delirium. Cox proportional hazards models were used to examine the association between HRV parameters and postoperative events by adjusting for delirium and pre and postoperative information. A total of 71 patients, 39 without delirium and 32 with delirium, met the inclusion criteria. The incidence of death and readmission within 1 year was significantly higher in the delirium group and in the group with higher daytime HF (high frequency power) and r-MSSD (square root of the squared mean of the difference of successive NN intervals), parameters of the parasympathetic nervous system activity (PNSA), than that in other groups. Furthermore, the delirium group had significantly higher HF and r-MSSD than the nondelirium group. Even after adjusting for confounding factors in the multivariate analysis, a trend of higher daytime HF and r-MSSD was observed, indicating a significant effect on the occurrence of combined events within 1 year of discharge. ICU delirium has been associated with higher daytime HF and r-MSSD, parameters of PNSA. ICU delirium was a prognostic factor, and increased daytime PNSA may worsen the prognosis of elderly patients after cardiovascular surgery.