The aim of this systematic review is to evaluate the available evidence, from randomized clinical trials (RCTs), of acupuncture for treating patients with RA. Systematic searches were conducted on 17 databases up to April 2008 without the language restriction. All RCTs of acupuncture, with or without electrical stimulation or moxibustion, for patients with RA were considered for inclusion. A total of 236 potentially relevant studies were identified and eight RCTs were included. Four RCTs compared the effects of manual or electro-acupuncture with penetrating or non-penetrating sham acupuncture and failed to show specific effects of acupuncture on pain [n = 88; weighted mean differences (WMD), 10 cm VAS -0.46; 95% CI -1.70, 0.77; P = 0.46; heterogeneity: tau(2) = 0.19; chi(2) = 2.38; P = 0.30; I (2) = 16%] or other outcome measures. One RCT compared manual acupuncture with indomethacin and suggested favourable effects of acupuncture in terms of total response rate. Three RCTs tested acupuncture combined with moxibustion, vs conventional drugs and failed to show that acupuncture plus moxibustion was superior to conventional drugs in terms of response rate (n = 345; RR 1.12; 95% CI 0.99, 1.28; P = 0.08; heterogeneity: tau(2) = 0.00; chi(2) = 1.34; P = 0.51; I(2) = 0%), pain reduction (n = 105; WMD, 10 cm VAS 1.53; 95% CI -0.57, 3.63; P = 0.15; heterogeneity: tau(2) = 1.18; chi(2) = 1.81; P = 0.18; I(2) = 45%) or joint swelling index (n = 105; WMD, 10 cm VAS 0.25; 95% CI -1.31, 1.82; P = 0.75; heterogeneity: tau(2) = 0.18;chi(2) = 1.14; P = 0.28; I(2) = 13%). In conclusion, penetrating or non-penetrating sham-controlled RCTs failed to show specific effects of acupuncture for pain control in patients with RA. More rigorous research seems to be warranted.