Background: Our prospective pilot study of acupuncture affecting insulin sensitivity on polycystic ovary syndrome (PCOS) combined with insulin resistance (IR) showed that acupuncture had a significant effect on improving the insulin sensitivity of PCOS. But there is still no randomized controlled trial to determine the effect of acupuncture on the insulin sensitivity in women with PCOS and IR. In this article, we present the protocol of a randomized controlled trial to compare the effect of true acupuncture on the insulin sensitivity of these patients compared with metformin and sham acupuncture. Acupuncture may be an effective therapeutic alternative that is superior to metformin and sham acupuncture in improving the insulin sensitivity of PCOS combined with IR. Methods: This study is a multi-center, controlled, double-blind, and randomized clinical trial aiming to evaluate the effect of acupuncture on the insulin sensitivity in PCOS combined with IR. In total 342 patients diagnosed with PCOS and IR will be enrolled. Participants will be randomized to one of the three groups: (1) true acupuncture + metformin placebo; (2) sham acupuncture + metformin, and (3) sham acupuncture + metformin placebo. Participants and assessors will be blinded. The acupuncture intervention will be given 3 days per week for a total of 48 treatment sessions during 4 months. Metformin (0.5 g per pill) or placebo will be given, three times per day, and for 4 months. Primary outcome measures are changes in homeostasis model assessment of insulin resistance (HOMA-IR) and improvement rate of HOMA-IR by oral glucose tolerance test (OGTT) and insulin releasing test (Ins). Secondary outcome measures are homeostasis model assessment-beta (HOMA-beta), area under the curve for glucose and insulin, frequency of regular menstrual cycles and ovulation, body composition, metabolic profile, hormonal profile, questionnaires, side effect profile, and expectation and credibility of treatment. Outcome measures are collected at baseline, at the end of treatments, and 3 months after the last acupuncture treatment. On completion of the screening visit, randomization will be conducted using a central randomization system. Discussion: This study will investigate the effects of acupuncture on the insulin sensitivity of PCOS and IR women compared with metformin and sham acupuncture. We will test whether true acupuncture with needles placed in skeletal muscles and stimulated manually and by electrical stimulation is more effective than metformin and sham acupuncture with superficial needle placement with no manual or electrical stimulation in improving the insulin sensitivity in PCOS women with IR.