BACKGROUND: Heart rate variability (HRV) reflects the balance between cardiac parasympathetic and sympathetic autonomic influences. Reduced HRV has adverse prognostic implications. The time course for changes in HRV over prolonged periods of time and the influence of an acute coronary event on HRV are not well established. METHODS: Heart rate variability was assessed in patients with chronic stable angina pectoris, who were followed for 3 years within the Angina Prognosis Study in Stockholm. Patients who suffered an acute myocardial infarction after the study were re-examined after this event. We assessed HRV by the simple geometric method differential index, and traditional time- and frequency-domain measurements of HRV. RESULTS: The differential index was essentially unchanged during the study (i.e. the ratio month 36/month 1 was 1.00 +/- 0.06, n = 261). Also most other time and frequency indices of HRV (SDNN, r-MSSD, SDNNIDX, total power, and VLF, LF, HF respectively; n = 63) remained largely unchanged; pNN50 and LF/HF were, however, less reproducible. In 21 patients with a subsequent acute myocardial infarction, SDNN, SDNNIDX, total power, LF and LF/HF were reduced following the event, whereas differential index, pNN50 and HF remained unchanged. CONCLUSIONS: Differential index and other indices of HRV are stable and reproducible in patients with chronic stable angina pectoris. High-frequency HRV (reflecting cardiac parasympathetic activity) and the differential index changed little following an acute coronary event, and may be suitable for predictions of the future risk of sudden death even in the presence of a recent acute coronary event.