OBJECTIVES: To investigate if, in a healthy randomly-selected population of 35-year-old men and women, there already is a relation between decreased heart rate variability and conventional risk factors for cardiovascular disease. BACKGROUND: Analysis of heart rate variability (HRV) has been used for estimating tonic autonomic activity. HRV is reduced in patients with various cardiovascular diseases. If a decreased HRV is a late phenomenon in the progression of cardiovascular disease, or if it parallels or even precedes manifest disease is unknown. DESIGN: Spectral analysis of HRV was made from 24-hour ambulatory electrocardiograms in a randomised population of healthy 35-year-old men (n = 63) and women (n = 70). The different spectral indices of HRV were analysed against gender, leucocyte count (previously described as an independent risk factor for cardiovascular disease), blood lipoproteins, smoking, heredity, body mass index (BMI) and systolic blood pressure. SETTING: A research centre of general medicine and a university hospital. RESULTS: Mean heart rate was lower, total power (TP), very low frequency power (VLF) and low frequency power (LF) were higher in men than in women. In women TP, VLF and LF were negatively correlated to BMI, smoking, triglycerides and positively correlated to HDL cholesterol. TP and VLF were also negatively correlated to risk factor score. High frequency power (HF), a marker of parasympathetic activity, was positively related to HDL cholesterol. In men, at daytime, TP, VLF, LF and HF were negatively correlated to leucocyte count. TP, VLF and LF were also negatively correlated to triglycerides and VLF also to risk factor score. CONCLUSIONS: There are correlations between HRV and known risk factors for cardiovascular disease already in a healthy 35-year-old population. A novel observation is the relationship in men between leucocyte count and heart rate variability.