Deep pain thresholds of the tender points in man were measured by pulse algometry. In this technique an insulated acupuncture needle (180 microns in diameter, about 400 k omega at 1 kHz) was used as a cathodal monopolar stimulating electrode. The stimulus consisted of a short train (three 1 msec duration pulses at 500 Hz) repeated at 2 Hz, and the current intensity was increased at a constant rate by a command voltage supplied to the stimulator. The stimulus current was monitored on an oscilloscope, and the integrated current was recorded simultaneously on a chart recorder. The tender points were detected by careful palpation, a pressure algometer being used to quantify threshold intensities. The needle electrode was inserted into the same points and deep pain thresholds were measured every 1-2 mm in depth. The deep pain thresholds at the depth where the experimenter felt some stiffness against the needle insertion (what we have termed needling stiffness) were 0.9 +/- 0.1 mA (mean +/- S.E.M., n = 14). Threshold values determined by repetitive measurements at the same position at intervals of 5 min for 6 successive trials were very similar. The stimulus evoked sensation was a dull-heavy deep pain. The pain thresholds changed markedly as the depth of the electrode tip was changed by 1-2 mm (P less than 0.01, t test). The minimum pain threshold under the tender point occurred at the depth of the needling stiffness. Observations with an ultrasonic image scanner suggested that the needle tip was located on or near the fascia when the experimenter felt the needling stiffness.(ABSTRACT TRUNCATED AT 250 WORDS)