Endometriosis is a condition in which tissue similar to the lining of the womb grows outside it, most often on the pelvic organs, and its commonest symptoms are painful periods (dysmenorrhoea), chronic pelvic pain and pain during sex (dyspareunia). The 2011 Cochrane review of acupuncture for pain in endometriosis found one trial that met its criteria and judged the evidence limited. Reviews published since 2017 pool between 6 and 42 trials and report lower pain scores with acupuncture, most of them against drugs or herbal medicine rather than a sham. In the two blinded sham-controlled trials, the differences in pain narrowed or disappeared over follow-up. This page reports what each source found and what each compared acupuncture with.
The Cochrane review
The review, by Xiaoshu Zhu, Kate Hamilton and Ewan McNicol, searched English and Chinese databases to 2010 for randomised, blinded trials in women whose endometriosis had been confirmed by laparoscopy, the keyhole surgery that is the usual way to diagnose it. Of 24 studies of acupuncture for endometriosis, one trial, with 67 participants, met every criterion. It compared auricular (ear) acupuncture with Chinese herbal medicine. Dysmenorrhoea scores were 4.81 points lower with ear acupuncture on a 15-point scale taken from a Chinese guideline for research on new Chinese medicines, and 91.9% of the ear acupuncture group against 60% of the herbal group met the trial's definition of an effective result. The improvement rate differed only among women with severe dysmenorrhoea. No data were available on quality of life, pregnancy, adverse effects or recurrence. The authors called the evidence limited and asked for well-designed, double-blinded trials. The review has not been updated since 2011: the DOI registry holds no later version, and PubMed records no update or withdrawal.
Later reviews
The reviews below differ in what they let in. Some pool trials against sham acupuncture, some against hormonal drugs or Chinese herbal medicine, and some count acupuncture added to drugs. A standardised mean difference (SMD) or Hedges' g expresses an effect in standard deviations rather than scale points, so trials that used different pain scales can be pooled.
| Review | Searched to | Trials (participants) | Comparisons | What it reported |
|---|---|---|---|---|
| Xu and colleagues, 2017 | December 2016 | 10 (589) | Mostly drugs and herbal medicine; one placebo-controlled pilot | Pain 1.36 points lower on a 0 to 10 scale than in control groups; arms of 8 to 36 patients |
| Wang and colleagues, 2023 | April 2022 | 15 (1,018) | Sham, conventional drugs, acupuncture added to drugs | Comparisons and outcomes too varied to pool; against sham, single studies reported lower period, pelvic and intercourse pain |
| Giese and colleagues, 2023 | March 2023 | 6 (331) | Non-specific (sham) acupuncture; usual care | Against sham, lower overall pelvic pain (g 1.54, 3 trials, 231 people) and period pain (g 1.67, 1 trial, 106 people); against usual care, lower period pain in 1 trial of 19 people |
| Chen and colleagues, 2024 | December 2022 | 14 (793) | Placebo, Chinese herbal medicine, Western drugs, pooled together | Lower pain severity (SMD −1.10) and lower serum CA-125, a blood marker raised in endometriosis |
| Yang and colleagues, 2025 | March 2025 | 9 (535) | Acupuncture alone against any treatment without acupuncture | Pain 1.67 points lower on a visual analogue scale; no difference in CA-125 |
Only the review by Nicola Giese, Kwan Kyung Kwon and Mike Armour rated its evidence with GRADE, the standard system for grading how far a body of evidence can be trusted. It rated the sham comparison for overall pelvic pain low, the single sham-controlled trial of period pain moderate, and the usual-care comparison very low. The review by Wang and colleagues graded its single-study sham results moderate to high and concluded that the evidence on the overall effectiveness of acupuncture and moxibustion for endometriosis was insufficient. Chen and colleagues, and Yang and colleagues, both name the design and quality of the included trials as limits on their findings.
Two network meta-analyses rank acupuncture-related methods against each other through shared comparators. Li and colleagues (2024) included 42 trials and 3,635 participants across six methods, and ranked acupoint catgut embedding, in which absorbable thread is left under the skin at a point, first for lowering pain. Su and colleagues (2025) included 23 trials and 1,545 patients of acupuncture-related methods combined with drugs, and ranked ear electroacupuncture first for overall symptoms. Both report their results against drug treatment rather than against a sham, and both called for higher-quality trials.
Trials that shaped the evidence
A 2008 trial by Peter Wayne and colleagues in Boston randomised 18 young women aged 13 to 22 with laparoscopically diagnosed endometriosis to Japanese-style acupuncture or a sham, 16 treatments over eight weeks. Its pain outcome was pelvic pain outside periods and intercourse. After four weeks that pain had fallen by an average 4.8 points on an 11-point scale in the acupuncture group (9 women) against 1.4 points in the sham group (5 women). Pain in the acupuncture group stayed lower through six months, but after four weeks the difference between the groups was no longer statistically significant. The trial was designed to test whether a larger one was feasible.
A 2010 Austrian crossover trial by Katharina Rubi-Klein and colleagues gave 101 women aged 20 to 40 two series of 10 sessions, each over five weeks, on top of their conventional treatment. One group received acupuncture at specific points first and non-specific acupuncture second; the other received them in the reverse order, with at least two menstrual cycles between the series. Of the 83 women who finished, the group given specific points first reported significantly lower pain after the first series, and the other group only after crossing over.
The largest sham-controlled trial is a 2023
multicentre trial by Li and colleagues at four hospitals in Jiangxi and Hainan provinces, China.
It randomised 106 women aged 20 to 40 to acupuncture or sham acupuncture, three 30-minute sessions a
week for 12 weeks, starting a week before each expected period, then followed them for 12 more
weeks. Its published protocol names five
points, CV04 (Guanyuan),
SP06 (Sanyinjiao),
LV03 (Taichong),
KD06 (Zhaohai) and
ST30 (Qichong), against a sham of shallow needles on the
shoulder and upper arm away from any acupuncture point, without stimulation. At 12 weeks the
acupuncture group reported a greater fall in dysmenorrhoea scores, shorter periods of pain, and
greater improvement in mood and quality-of-life scores. By week 24 the differences in period pain
and questionnaire scores were gone. Pelvic pain outside periods and pain during sex did not differ
between the groups, and the trial recorded no serious adverse events.
A 2021 Australian feasibility trial by Mike Armour and colleagues in Sydney compared 16 sessions over eight weeks plus usual care with usual care alone. Of 29 eligible women, 19 completed it: 14% withdrew from the acupuncture group and 53% from usual care. Median period and non-period pain fell by about 2 points in the acupuncture group only, and adverse events, mostly mild, occurred in 6.7%. The authors concluded that usual care alone was not an acceptable control for a larger trial.
What guidance says
The European Society of Human Reproduction and Embryology (ESHRE) endometriosis guideline, issued on 2 February 2022 and, on ESHRE's guideline page, still its latest version, asks clinicians to discuss non-medical strategies for quality of life and psychological well-being. It makes no recommendation for any specific non-medical intervention, acupuncture, electrotherapy and Chinese medicine among them, because their potential benefits and harms are unclear. It reaches the same conclusion for fertility.
NICE's guideline NG73, Endometriosis: diagnosis and management, was published in September 2017 and last updated in November 2024, when its diagnosis recommendations were revised. Under non-pharmacological management it has one recommendation, unchanged since 2017: to advise women that the available evidence does not support traditional Chinese medicine, other Chinese herbal medicines or supplements for treating endometriosis. It does not mention acupuncture. Clinical guidelines sets out what other guideline bodies say, condition by condition.
What the reviews could not settle
Most pooled trials compared acupuncture with drugs or herbal medicine, and many were small and unblinded, so they cannot separate the effect of needling from the attention and expectation that come with a course of treatment. The sham-controlled evidence rests on a handful of trials, and the largest found its differences gone 12 weeks after treatment stopped. Sham acupuncture explains why no sham is inert. The reviews had too few comparable trials to say which points, styles or numbers of sessions matter, and the largest sham-controlled trial found no difference in pain during sex. The Cochrane review found no trial data on quality of life, adverse effects or recurrence, and Wang and colleagues still judged the evidence on overall effectiveness insufficient in 2023.
Period pain reports the evidence for primary dysmenorrhoea, where no underlying disease is found; Chronic pain covers the individual patient data meta-analyses across chronic pain conditions; and Uterine fibroids and Polycystic ovary syndrome report the trials in two other gynaecological conditions. AcuiQ’s endometriosis page lists the protocols individual studies prescribed, each with its citation, and Reading a protocol explains each field of an entry.