The role of spinal manipulation in the relief of pain is becoming clearer and more demonstrable as time passes. One approach to this study is the effect of manipulation on the neurochemical mechanisms of antinociception. Chief among these is beta-endorphin, which has been found to produce a wide range of beneficial effects, especially analgesia. The intent of our study was to demonstrate the effect of spinal manipulation on plasma beta-endorphin levels. Three groups of male subjects were randomly created: the experimental, sham and control groups. All three groups were screened for symptomatology, present use of medications and the present use of innocuous stimulants, such as nicotine and caffeine. A standard protocol involving a 20-min pretest resting period, an intervention and a 40-min test period ensued. The experimental group received a manipulation in the region of the cervical spine; the placebo group received a sham maneuver with no dynamic thrust; the control group received no intervention. Samples were taken by venipuncture at -20, -5, +5, +10 and +30 min. The data were analyzed by repeated measures analysis of variance and by Scheffe's post-hoc multiple comparison tests. Plasma beta-endorphin levels were assessed by radioimmune assay technique (according to the method described by Harber and Sutton in 1984). The results of our study demonstrated a small, but statistically significant, increase in serum beta-endorphin levels in the experimental group at the 5-min postintervention point. The levels in the placebo and control groups demonstrated a steady decrease that was distinct from the response in the experimental group.(ABSTRACT TRUNCATED AT 250 WORDS)