Objectives: To investigate the contribution of two principal features that underlie traditional Chinese acupuncture: site specificity and application of needle manipulation. Design: Thirteen volunteers completed a randomised, dual blind (subject and assessor) repeated measures study involving five interventions. Pressure pain threshold (PPT) was measured with an algometer, before and after intervention at 10 sites (acupoints and nonacupoints) across the body. Interventions: Deep needling, with or without manual needle rotation, applied to the acupoint Large Intestine 4 (LI4) or to a nonacupoint located on the medial side of the second metacarpal. Inactive laser to LI4 was used as a control. All interventions were administered for 21 min. Main outcome measures: Percentage change in PPT from preintervention baseline at the 10 sites during the 18 min immediately following intervention. Results: Statistically significant increases from preintervention PPT means were obtained at all 10 sites following needling of LI4 with manipulation compared with one site after needling LI4 without manipulation. Needling the nonacupoint led to statistically significant increases at six sites when manipulation was present compared with none in the absence of manipulation. No significant changes in mean PPT followed inactive laser. Needling LI4 with manipulation produced mean increases that were statistically significantly greater than those for the other interventions with one exception: needling the nonacupoint with manipulation was as effective as needling LI4 with manipulation at one measurement site only. Conclusions: Both manipulation and site of needling contributed significantly to the elevation of PPT following acupuncture. Distribution of effects on PPT did not support either neural segmental or Traditional Chinese Medicine channel theories. Psychological and physiological nonspecific effects appeared to play a minimal role in changes to PPT.