The aim of the present study was to determine if acupuncture stimulation inhibits sympathetic nerve activity in humans. Multiunit efferent postganglionic sympathetic activity was recorded with a tungsten microelectrode inserted in a muscle fascicle of the peroneal nerve. Mean arterial pressure, heart rate and skin blood flow were also monitored. Pain thresholds were measured by electrical tooth pain stimulation. After a 30 min rest, acupuncture needles were inserted bilaterally into the Li 11 and the Li 4 acupuncture points, and manipulated until ‘chi’ cramp-like sensation was reported. Electrical stimulation (2 Hz, 0.6–0.8 ms duration, maximal tolerated stimulation without discomfort) was delivered for 30 min and the physiological recordings were continued for 90 min after the end of acupuncture. In a placebo control experiment, the same procedure was followed, except that acupuncture needles were inserted subcutaneously and no manipulation or stimulation was given. The stimulator delivered pulses to an unconnected channel, hence, the same audiovisual stimuli were experienced as with acupuncture, and care was taken to ask the same questions about sensations in the placebo and the acupuncture groups. Electroacupuncture produced an increase in pain threshold which was paralleled by a transient increase in muscle sympathetic nerve activity. During acupuncture, there was a small increase in heart rate and mean arterial pressure, but there was no post-acupuncture hypotension. The placebo control procedure did not change pain threshold or sympathetic nerve traffic. The findings suggest that electroacupuncture produces moderate hypoalgesia in humans paralleled by a significant increase in muscle sympathetic nerve activity.