Fire Cupping Leaves Real Marks. The Evidence It Does Much Else Is Thin.

Educational journalism, not medical advice. Every claim here is checked against its cited sources by editor Tim Bunce — a health writer, not a physician. It isn’t specific to your situation: for health decisions, talk to your own clinician. How we work →

What cupping claims

Cupping — including Chinese fire cupping, dry cupping, and wet (‘hijama’) cupping — uses suction to pull skin and tissue into a cup, often explained as ‘improving blood and qi flow,’ ‘releasing fascia,’ or ‘drawing out toxins.’ Two of those claims are testable, and one is just wrong. Let’s start with the marks.

The marks are bruises, not toxins

The colored circles are extravasated blood — suction ruptures tiny capillaries and blood pools under the skin, then reabsorbs over days (red → purple → green-yellow, exactly like a bruise). They reflect suction pressure and how fragile your vessels are, not how many ‘toxins’ came out. A darker mark is not a deeper cleanse.

Does it help pain? Short-term, on weak evidence

Many trials report short-term reductions in neck and back pain Kim 2018 Kim 2011, and a large 2025 update echoed a pain benefit — but flagged that all 72 trials were at high risk of bias Wang 2025. The foundational review of 135 trials found 84% were high risk of bias and none were low-risk Cao 2012. So ‘it helped in studies’ comes with a heavy asterisk.

The placebo problem

Here’s the tell. In the best-controlled trial — a randomised study in fibromyalgia — cupping beat usual care but did not beat sham cupping Lauche 2016. That pattern says the benefit is largely nonspecific: expectation, attention, and hands-on contact. Cupping is also nearly impossible to blind — the marks give it away — so expectation contaminates almost every study.

Athletes and recovery

Reviews in athletes report possible benefits for pain, range of motion, and soreness markers, but on trials with unclear-to-high risk of bias, and most reported no safety data Bridgett 2018; broader musculoskeletal/sports reviews reach the same ‘maybe, but weak and inconsistent’ verdict Mohamed 2023. It is not a substitute for established rehab.

Safety, said plainly

This is the part that matters most. Fire/dry cupping: the main risk is burns — an Australian burns-registry study found most cupping burns were flame burns involving an accelerant, usually at home Seifman 2017; lesser risks are skin irritation, soreness, and lingering marks Kim 2014. Wet cupping (which cuts the skin) carries real bloodborne-infection risk if equipment isn’t sterile, including rare but documented serious and fatal infections Alajmi 2021. Authorities call the overall evidence insufficient NCCIH. Avoid if you have a bleeding disorder, take blood thinners, have fragile skin or eczema, or are pregnant — and don’t let it delay real care for a real injury.

The bottom line

Cupping is best understood as a low-cost comfort ritual that might take the edge off muscle soreness, largely through placebo and hands-on attention — not a proven treatment and not a detox. If you enjoy it and choose a trained, hygienic practitioner, the dry/fire version is usually low-harm. But the marks are bruises, the evidence is thin, and the fire and blade are where the real risk lives.

• The marks are bruising (pooled blood), not toxins — darker isn’t ‘more out.’
• Short-term pain relief appears in trials, but they’re low-quality and hard to blind.
• Didn’t beat sham in the best-controlled test — consistent with placebo.
• Real risks: burns (fire) and infection (wet cupping). Not a substitute for medical care.

How cupping is supposed to work — and how thin those theories are

If the marks are just bruises, why would cupping change anything at all? Practitioners and researchers have floated several physiological ideas, but it is worth being upfront from the start: a 2019 review in the Journal of Traditional and Complementary Medicine that catalogued them concluded that "the exact mode of action of cupping to reduce pain is not well understood," and that "no single theory could explain the mechanisms of action" Al-Bedah 2019. In other words, these are plausible-sounding hypotheses, not settled biology.

The most-cited idea borrows from pain science. The pain-gate theory holds that strong touch and pressure signals can crowd out pain signals before they reach the brain — the same reason rubbing a banged elbow helps. A related idea, diffuse noxious inhibitory controls (DNIC, sometimes called "pain inhibits pain"), proposes that a new, mild painful stimulus in one spot — the pull of the cup, say — triggers the brainstem to dial down pain perception elsewhere Al-Bedah 2019. A separate strand is the nitric oxide theory: the suction creates local low-oxygen conditions that may prompt blood vessels to release nitric oxide, a molecule that widens vessels and briefly boosts skin blood flow Al-Bedah 2019.

Two cautions matter here. First, even if cupping does nudge skin blood flow or trigger a pain-gate response, those mechanisms would explain a short, modest, local effect — not the sweeping disease cures sometimes advertised. Second, demonstrating a biological signal (more blood flow, a measurable change in skin) is not the same as demonstrating a clinical benefit; plenty of interventions move a lab number without making patients feel better. The honest summary is that cupping has biologically reasonable ways it might ease soreness for a little while, and no proven mechanism for anything bigger.

The methods, and what a typical session looks like

"Cupping" is an umbrella term, and the variety matters for both how it might work and how risky it is. The main forms, as catalogued in the NIH's StatPearls clinical reference, are dry cupping (cups create suction on intact skin, with no cutting); wet cupping, also called hijama (the skin is lightly scratched or incised first, so the cup draws out a little blood and fluid); flash cupping (cups are quickly applied and removed in succession); and running or massage cupping (oil is applied and the cup is slid across the skin) StatPearls 2024. "Fire cupping" simply refers to how the vacuum is made — a flame briefly heats the air inside a glass cup before it is placed on the skin — and can be dry or wet. The flame is gone before the cup touches you; it is the cooling air that creates the pull.

On dose, there is no rigorously validated protocol, but the practical convention is modest. StatPearls notes that "cups should be left on the skin for a maximum of 5 to 10 minutes" per session, and the resulting marks "typically fade … within 1 to 10 days" StatPearls 2024. This is useful context for reading the research: the trials that report short-term pain relief generally used a handful of sessions over a few weeks Wang 2025, so claims that one session delivers lasting change, or that longer-is-better, are not backed by the evidence. If you try it, a session that leaves marks lasting much longer than about ten days, or that blisters, suggests the suction was too strong or held too long.

Who should think twice before cupping

Because dry cupping is low-tech, it is easy to assume it is risk-free for everyone. It is not. The same StatPearls reference lists several situations where cupping is contraindicated or should be approached with caution: people taking anticoagulant ("blood thinner") medication, those with haemophilia or other bleeding disorders, and people with cancer or organ failure; it also advises against cupping over broken, inflamed, or infected skin, over varicose veins, or directly over major vessels, and notes it "is not recommended for geriatric, pediatric, pregnant, or menstruating patients" StatPearls 2024. The logic is straightforward: cupping deliberately ruptures small blood vessels, so anything that impairs clotting or skin integrity raises the odds of large bruises, blood collections under the skin (haematomas), or — with wet cupping — bleeding and infection.

These cautions are general guidance, not a personal risk assessment. The same review stresses that taking a careful medical history beforehand is essential to minimise complications StatPearls 2024. If you are pregnant, have a bleeding or clotting condition, take blood thinners or antiplatelet drugs, have diabetes or fragile skin, or are managing a serious illness, talk to your own clinician before trying cupping — and treat it as a comfort add-on, never a replacement for evidence-based care. Wet cupping in particular carries infection risk and should only be considered, if at all, with sterile single-use equipment and a trained practitioner.

Why the evidence stays weak — the blinding problem

It is fair to ask why, after thousands of years and hundreds of trials, the verdict is still "maybe, weakly." A big part of the answer is that cupping is genuinely hard to study well. The core difficulty is blinding: you cannot easily hide from a participant whether a cup is suctioning their skin, and you certainly cannot hide the circular marks afterward. As the authors of a 2025 meta-analysis put it, "without placebo control they were impossible to blind the clinicians and participants" — and they rated all 72 of their included trials as high risk of bias Wang 2025. That echoes an earlier, larger review in which 84% of 135 trials were high risk of bias and not one met low-risk criteria Cao 2012.

Researchers have tried to solve the blinding problem with sham cupping. One validated approach drills a tiny hole in the cup so suction cannot build, letting it act as a placebo Lee 2010. But sham devices are imperfect: in the best-controlled fibromyalgia trial, most participants still correctly guessed which treatment was real, and crucially, genuine cupping did not beat the sham Lauche 2016. When a "fake" version that creates no real suction works about as well as the real thing, the most parsimonious reading is that much of the benefit is expectation, attention, and hands-on contact rather than the suction itself.

Two further factors should temper enthusiasm. First, publication bias: the 2025 review found that across its analyses the data "showed asymmetry and suggested potential publication bias," meaning small studies with disappointing results may never have been published, inflating the apparent benefit Wang 2025. Second, the evidence base is geographically lopsided — 62 of those 72 trials were conducted in China and published in Chinese Wang 2025, a literature where independent reviews have flagged unusually high rates of positive findings. None of this proves cupping does nothing. It means the studies that look most encouraging are also the ones least able to rule out placebo, reporting bias, and chance — which is exactly why careful reviewers keep landing on "low-cost, low-harm, largely nonspecific," not "proven treatment."

References