Cupping sets cups on the skin and lowers the pressure inside them, so that skin and the tissue beneath are drawn up into the cup. The US Council of Colleges of Acupuncture and Oriental Medicine’s Clean Needle Technique Manual (seventh edition, revised 2017) divides it into fire cupping and suction cupping, both forms of dry cupping, and wet cupping, in which the skin is punctured first so that the cup draws blood. It records the practice in East Asian medicine, in the traditional medicine of the Gulf Arab countries as hijamah, in Europe, and in Greek, Roman and Egyptian medicine; a 2015 review by Pratiksha Mehta and Vividha Dhapte adds that the Greek physician Galen described cups of glass, horn and brass.
Cups and suction
In fire cupping, the manual says, a ball of burning cotton or a lit alcohol swab is held briefly inside a glass cup to heat the air, which forms a partial vacuum as it cools once the cup is on the skin. Suction cupping uses plastic or silicone cups with a valve, from which a hand pump draws out air. In both, cups usually stay on for 2 to 10 minutes and sometimes up to 20. The STRICTOC reporting standard for cupping trials, published in 2020 by Xuan Zhang and colleagues as an extension of the CONSORT statement, asks trials to name the cup material (plastic, glass, rubber, bamboo, ceramic, metal or silicone), its size, and how the suction was made and how strong it was. One trial it quotes used a device that pulsed the pressure between −150 and −350 millibars every 2 seconds.
Styles
| Style | What is done |
|---|---|
| Retained (stationary) | Cups are placed and left for a set time |
| Moving or sliding | A cup is moved over a lubricated area, along a channel or a muscle; one trial STRICTOC quotes moved cups up and down both sides of the spine |
| Flash or empty | Cups are removed as soon as the suction takes, and may be reapplied many times; STRICTOC asks trials to report how many times |
| Wet (hijama) | The skin is punctured with a lancet or plum-blossom needles, then cupped; STRICTOC describes a few drops of blood being drawn into the cup |
| Needle cupping | A cup is placed over an inserted acupuncture needle |
STRICTOC lists further Chinese styles, among them medicated cupping over herbs and water cupping, and notes that in the Middle East and Europe dry and wet cupping are the main forms. The Clean Needle Technique Manual warns that a cup over a needle can push it deeper or lift the tissue around it, which over the chest raises the risk of a collapsed lung; Needling over the chest covers that risk.
The marks
The round marks cupping leaves are small bleeds under the skin. The Clean Needle Technique Manual describes them as petechiae (pinpoint bleeds) and ecchymosis (bruising) in round patches, produced on purpose, which usually fade within a few days to 2 weeks. It notes that people unfamiliar with cupping have mistaken them for harm or for child abuse, that no study has set how long cups should stay on or how strong the suction should be, and that too much of either leaves bruising and darkened skin that last longer. One trial STRICTOC quotes removed the cups after 10 to 20 minutes according to the colour of the marks, from slightly rose to dark pink.
A 2017 review of the effects of suction on skin by Dennis Lowe describes how the marks form: the negative pressure stretches the skin and widens its capillaries, blood flow rises, and capillaries eventually rupture into the tissue. Scavenger cells then clear the escaped red cells, and the enzyme they use to break down the haem, haem oxygenase-1, produces carbon monoxide, biliverdin and bilirubin. Lowe goes on to propose that these breakdown products, which have anti-inflammatory effects in other settings, explain some of what cupping is used for; the review presents that as a hypothesis.
What reviews and trials found
| Source | What it covered | What it found |
|---|---|---|
| Cao, Li and Liu, PLoS One, 2012 | 135 randomised trials from 1992 to 2010, 132 of them in Chinese; wet cupping in 78 | None at low risk of bias and 84% at high risk. Cupping added to other treatment beat that treatment alone on the top grade of the trials’ four-grade efficacy scales for shingles, facial paralysis, acne and neck spondylosis; no serious adverse effects reported |
| Zhang and colleagues, Chinese Medicine, 2017 | 23 trials, 2,845 participants, 12 pain conditions, cupping against acupuncture | Similar results for the two in every condition; all trials of poor methodological quality, and a trial sequential analysis for neck spondylosis found the data too few for a firm conclusion |
| Jia and colleagues, BMJ Open, 2025 | 10 trials, 656 people with chronic musculoskeletal pain; controls were sham cupping in 3, a waiting list in 5 and rest in 2 | Lower pain straight after treatment (standardised mean difference −1.17, 95% CI −1.93 to −0.42, moderate certainty); no improvement in disability or mental health |
| Jenkins and colleagues, JOSPT Open, 2026 | 5 trials, 281 people with musculoskeletal pain, dry cupping against placebo cupping; literature searched to June 2025 | No clear difference from placebo at 1 week (mean difference −9.9 on a 0 to 100 pain scale, 95% CI −30.5 to 10.7) or at 6 months; one trial favoured cupping at 4 weeks; very low certainty throughout, with 4 of the 5 trials at high risk of bias |
| Lauche and colleagues, Scientific Reports, 2016 | 141 people with fibromyalgia, randomised to cupping, sham cupping or usual care | Less pain at day 18 than usual care (difference −12.4, 95% CI −18.9 to −5.9) but not than sham (−3.0, −9.9 to 3.9); minor side effects only |
Most of the trials Jia and colleagues pooled compared cupping with no cupping, and the effect they measured was the one that appears straight after a session, because their trials recorded nothing later. Lauche’s trial, which had both kinds of control, found the effect over usual care and none over sham, and its authors concluded that cupping could not be recommended for fibromyalgia at that time. Jenkins and colleagues, who pooled only trials with a placebo-cupping control, rated the effect on pain very uncertain at every time point and advised trying treatments with firmer evidence first. Cao and colleagues note that the four-grade efficacy scales most of their trials used have no internationally recognised meaning. AcuiQ’s pages on shingles and fibromyalgia list the protocols individual studies prescribed, each with its citation.
Burns, blisters and infection
The Clean Needle Technique Manual, citing a review of 12 years of case reports by Xu and colleagues, says most adverse events from cupping were minor, chiefly raised scars, burns and blisters, and that the burns come from fire cupping: a flame held too close to the rim heats the glass, or burning material is left inside the cup when it goes on the skin. Fluid blisters are not uncommon, and blood-filled ones are more likely in people taking anticoagulants or supplements that thin the blood. Wet cupping draws out blood and so carries a risk of passing on bloodborne infection, and the manual cites case reports of infection after it. Burns from moxibustion and cupping and Infection and clean needles cover both risks.
What the sources cannot tell you
Cao’s review counts many trials of low quality, most testing cupping on top of other treatment; the sham-controlled evidence is small; and the explanation of the marks does not show that the marks do anything for the condition being treated. A protocol entry on AcuiQ names cupping as its modality where the study used it, alongside the number of sessions and the length of the course; Reading a protocol explains each field.