The sympatho-vagal nerve interaction at the heart was studied by means of power spectrum analysis of heart rate variability in seven Caucasians (aged 27-35 years) in resting supine and sitting positions before and during 35 days of a sojourn at 5050 m above sea level (asl) and in six Sherpas (aged 22-30 years) at high altitude only. A high frequency peak (HF)-central frequency between 0.20 and 0.33 Hz, a low frequency peak (LF)-central frequency between 0.08 and 0.14 Hz, and a very low frequency component (< 0.05 Hz), no peak observed, were found in the power spectrum in both positions and independent of altitude. The peak powers, as a percentage of the total power, were affected by both body position and altitude. At sea level the change from a supine to a sitting position yielded a decrease in percentage HF from 25 (SEM 1.9)% to 6.2 (SEM 1.5)% (P < 0.05) and a significant increase in the ratio between LF and HF powers (LF:HF) from 1.7 (SEM 0.4) to 6.9 (SEM 1.6). At altitude compared to sea level in the supine position, percentage HF decreased from 25% to 10.9 (SEM 1.0)% (P < 0.05) and the LF:HF ratio increased from 1.7 to 4.8 (SEM 0.7) (P < 0.05). No changes occurred at altitude in the sitting position either in the peak powers or in the LF:HF ratio, but the central frequency of HF peak increased significantly from 0.25 (SEM 0.02) Hz to 0.32 (SEM 0.01) Hz. In the Sherpas comparable results to the Caucasians were found in both body positions. The high LF:HF ratios observed at altitude in both body positions and groups would suggest that hypoxia caused a shift of sympatho-vagal nerve interaction at rest toward a dominance of the sympathetic system, which was found at sea level only in the sitting position. An acclimatization period of 10 days higher than 2850 m asl and 1 month at 5050 m asl did not modify the interactions of the autonomic systems.