MDExperimentally induced cold pressor pain was used to: 1) compare acupuncture and placebo acupuncture; and 2) assess the effect of subject expectancy on treatment outcome. Seventy-two subjects were randomly assigned to three groups of 24 subjects. The first of two 60 sec cold pressor trials was identical for all groups and served as a base-line. The second trial was preceded by acupuncture for the first group, placebo acupuncture for the second, and no treatment for the third. Within each of these three groups, expectancy was manipulated prior to the second trial: eight of the 24 subjects were led to expect less pain on the second trial (high expectancy); eight were led to expect no change in pain (low expectancy); and eight were told that the effect of treatment could not be predicted (no expectancy). Numerical rating of pain intensity revealed no cases of dramatic pain reduction. Only subjects who received acupuncture treatment in conjunction with the high expectancy manipulation reported significant pain relief. Results are consistent with the view that acupuncture involves more than placebo factors, but the treatment appears to require a positive attitude of the recipient to potentiate its effect.The recent literature on pain control contains numerous studies of the effect of acupuncture on experimentally induced pain. Some investigators have reported reliable pain reduction with acupuncture (e.g., 1,3,7,8,11,16,22,25,26,28), but with two exceptions (2, 6) the effect has been small. Further, others have found acupuncture analgesia to be either ineffective or no more effective than a placebo treatment (e.g., 5, 9,10,12,19, 20, 24). It is clear that, at least for experimental pain, the effect of acupuncture analgesia is limited. However, debate continues on the underlying mechanisms of acupuncture analgesia; it is not clear whether specific physical properties of the acupuncture treatment procedure contribute to its effect or whether the effects of acupuncture analgesia are entirely psychological. Attempts to resolve this issue have met with a major methodological problem: the development of satisfactory double-blind procedures for the study of acupuncture analgesia. Such a procedure requires a placebo treatment which is indistinguishable from legitimate acupuncture by both the acupuncturist and the subject.The most commonly used placebo acupuncture procedure involves the insertion and stimulation of needles at points on the body which are close to the traditional acupuncture points under study but which are not themselves acupuncture points (e.g., 1,12,24,25,28).