Abstract Objective: Conventional treatment for RP is limited due to side effects, and complementary and alternative medicines (CAM) are widely used by the population. Our objective was to find an effective and well-tolerated CAM for the treatment of RP. Methods: Using MEDLINE, EMBASE and AMED, 20 randomized controlled trials (RCTs) were found and divided into nine treatment subcategories: acupuncture (n = 2 trials), anti-oxidants (n = 2), biofeedback (n = 5), essential fatty acids (n = 3), Ginkgo biloba (n = 1), L-arginine (n = 2), laser (n = 3), glucosaminoglycans (n = 1) and therapeutic gloves (n = 1). Trials in each subcategory were meta-analysed together. Results: Several categories did not have enough trials to do a meta-analysis and most trials were negative, of poor quality and done prior to 1990. Biofeedback was negative for a change in frequency, duration and severity of RP attacks, and actually favoured control (sham biofeedback; P < 0.02). The therapeutic glove favoured active treatment (P < 0.00001). Laser resulted in one less RP attack on average over 2 weeks vs sham [weighted mean difference (WMD) 1.18; 95% CI 1.06, 1.29], and a change in severity of attacks (WMD 1.98; 95% CI 1.57, 2.39; P < 0.05). No significant differences were found in the nutritional supplements that were studied. Conclusions: There is a need for well-designed trials of CAM in RP. The literature is inconclusive except that biofeedback does not work for RP, therapeutic gloves may improve RP (but results may not be generalizable due to single trial site and no intent-to-treat analysis) and laser may be effective but the improvement may not be clinically relevant.