Transient ischemia on Holter monitoring is a major determinant of outcome in unstable angina. In this study we investigated whether analysis of heart rate variability (HRV) may further improve the prognostic yield of Holter monitoring in this clinical setting. We performed 24-hour Holter monitoring in 75 patients with unstable angina (59 men, aged 62 +/- 9 years) within 12 hours of hospital admission. Number and duration of myocardial ischemic episodes, and both time domain and frequency domain HRV measures were obtained from Holter recordings. In-hospital major cardiac events (death or myocardial infarction) occurred in 7 patients (9%). Episodes of ST-segment depression on Holter monitoring were found in 6 of 7 patients (86%) with and in 26 of 68 patients (38%) without events (p <0.05). There were no differences between patients with or without events in both time domain (standard deviation [SD] of all normal RR intervals in the entire 24-hour electrocardiographic recording (SDNN), SD of the mean RR intervals for all 5-minute segments (SDANN-i), mean of SD of all RR intervals for all 5-minute segments (SDNN-i), percentage of differences between adjacent RR intervals >50 ms (pNN50), and square root of the mean squared differences of successive RR intervals) (RMSSD), and frequency domain (ultra low, very low, low, and high frequency) HRV indexes. However, the low-frequency/high-frequency (LF/HF) ratio was significantly higher in patients with cardiac events (2.12 +/- 1.4 vs 1.48 +/- 0.5, p = 0.01). Moreover, when considering only the 32 patients with myocardial ischemic episodes on Holter monitoring, the LF/HF ratio was again higher in the 6 patients with than the 26 patients without major cardiac events (2.45 +/- 1.5 vs 1.31 +/- 0.3, p <0.01). Multivariate logistic regression, including clinical and angiographic variables, showed that transient ischemia on Holter monitoring was the only independent determinant of outcome (odds ratio = 12.2, p = 0.03), with the LF/HF ratio being only slightly over statistical significance (odds ratio for 0.1 increments = 2.8, p = 0.08). Our data confirm that transient ischemia on Holter monitoring is a powerful predictor of cardiac events in unstable angina and indicates that an imbalance in cardiac autonomic tone toward a prevalence of sympathetic activity increases the risk of events in this group of patients.