BACKGROUND: Heart rate variability (HRV), a measure of cardiac autonomic balance seems to be linked to coronary artery disease (CAD). Impaired vagal input facilitates the generation of fatal arrhythmias and has a great impact on morbidity and mortality. The purpose of this study was to determine the relationship between perioperative HRV and the incidence of silent myocardial ischemia (SMI) and ventricular dysrhythmias in CAD-patients undergoing non-cardiac surgery. METHODS: 31 patients were studied by continuous Holter electrocardiography preoperatively and again on the evening before surgery until postoperative day 3. Three frequency and one time domain measures of HRV (TP, 0.01-1.00 Hz: total power of the amplitude spectral plot; LF, 0.04-0.15 Hz: low-frequency power; HF, 0.15-0.40 Hz: high-frequency power; MeanRR: mean of all coupling RR-intervals between normal beats) as well as ischemic events and ventricular couplets and runs were computed. RESULTS: Depending on the presence of ischemic episodes, each patient was assigned to either the no SMI-group (13 pts) or the SMI-group (18 pts). MeanRR, TP and LF significantly declined in both groups over time. The parasympathetically dominated index HF, however, only decreased in the SMI-group. Normalized HF power (HF/TP) even increased in the no SMI-group resulting in a postoperative decrease in LF/HF ratio. TP, LF, and HF inversely correlated with ischemia parameters whereas HF/TP and HF/LF ratio correlated with the number of ventricular couplets. Incidence and severity of SMI significantly increased after surgery. CONCLUSION: Postoperatively, a prolonged vagal withdrawal occurred in CAD-patients exhibiting perioperative SMI. Whether the increased incidence of SMI after surgery in conjunction with the observed parasympathetic derangement contributes to adverse cardiac outcome still has to be determined.