The published research
Ozone therapy
A summary of the peer-reviewed research on medical ozone — what has been tested, what was found, and where the evidence runs out. Every statement below is linked to its source study.
What the research shows
The most useful place to start with ozone is the proposed mechanism, because it explains both why it might do anything and why concentration matters so much. The standard account is that ozone works by producing a controlled, moderate oxidative stress. Moderate oxidative stress activates a transcription factor called Nrf2, which switches on the cell's own antioxidant machinery — superoxide dismutase, glutathione peroxidase, catalase, heme oxygenase-1. Severe oxidative stress does something different and unwanted: it activates NF-κB and drives inflammation. The authors of the foundational review put it plainly — the line between effectiveness and toxicity depends on the strength of the oxidative stress, in the same way that moderate exercise is good for you and excessive exercise is not 1. Dose is not a detail here. It is the whole argument.
The best-studied clinical use is the knee. A 2024 level I meta-analysis pooled randomised controlled trials comparing intra-articular ozone against hyaluronic acid — a mainstream, widely used joint injection — in 424 patients, and found no significant difference in pain scores at four to six months 2. An earlier systematic review and meta-analysis of five randomised trials in 428 patients reported that ozone outperformed placebo, while performing slightly below other active injections at three to six months, a difference that did not reach statistical significance. Adverse events in that pooled analysis occurred in 4% of the ozone group and 5.4% of the control group 3. Set against the broadest available comparison — a network meta-analysis of 79 randomised trials and 8,761 patients that ranked eleven different injectables — ozone is one credible option among several rather than the leader 4.
Wound care is the other area with real depth behind it. A systematic review in the International Wound Journal covering twelve studies found that, in diabetic foot ulcers, ozone alongside standard care was associated with markedly faster improvement in wound area and a substantially lower amputation rate. The same review was careful about what it did not find: no difference in the proportion of wounds that fully healed, and no difference in length of hospital stay. No adverse events attributable to ozone were reported across those studies 5. A separate 2024 systematic review and meta-analysis reached similar conclusions on healing rates and hospital stay, and was blunt that a considerable risk of bias sat in the underlying trials 6.
Systemic ozone — major autohemotherapy, the treatment most of our menu is about — has been tested in randomised trials, but small ones. In a 2024 pilot randomised controlled trial of 73 patients, 71% of those receiving ozone autohemotherapy alongside conventional care met the pre-defined response threshold, against 45% on conventional care alone, with measurable differences in symptom scores, tidal volume and six-minute walk distance 7. A second pilot randomised trial in 40 critically ill, mechanically ventilated patients reported more ventilator-free days and shorter intensive-care stays in the ozone group 8. These are genuine randomised results and they are also small, single-centre and preliminary. They are the beginning of an evidence base, not the end of one.
Ozone is also used as a local injection for low back pain, where a meta-analysis found that image-guided injection outperformed injection placed by anatomical landmarks alone 9 — a finding about technique rather than about ozone, and a reminder that how a procedure is done is part of whether it works.
Finally, safety. A 2026 scoping review searched nine databases for documented adverse events from systemic oxygen-ozone autohaemotherapy. Its conclusion was that the current evidence does not show a pattern of frequent serious harm when a standardised protocol is followed — and it was specific about what that protocol involves: mandatory G6PD screening, staying within recommended ozone concentrations, controlled reinfusion rates, cardiovascular evaluation in selected patients, and strict aseptic technique. The serious adverse events it did find in the literature were largely case reports, and haemolysis and renal failure were associated with excessive ozone concentrations 10. The G6PD test we require before your first session is not a formality; it is the first item on that list.
Where the evidence is thin
We would rather tell you this ourselves than have you find it somewhere else. These are the parts of the picture that do not favour the treatment, or where the research simply has not been done.
Ozone is not FDA-approved, and that is not a technicality
Under 21 CFR 801.415 the FDA classifies ozone as a toxic gas with no known useful medical application in specific, adjunctive or preventive therapy — a position it has held since 1976. No ozone generator has FDA approval for therapeutic use in the United States.
That is a regulatory fact and it sits alongside, not instead of, the research above. Both are true. We would rather you weigh them yourself than discover one of them later.
The systemic evidence is pilot-scale
The randomised trials of major autohemotherapy cited here enrolled 73 and 40 patients. That is enough to justify further study. It is not enough to establish a treatment, and anyone telling you otherwise is overselling it.
Risk of bias is a recurring criticism
Reviewers of the wound-care literature explicitly flagged considerable risk of bias in the underlying trials, and called for better-designed research. Much of the ozone literature is small, single-centre, and not blinded.
The UV light add-on does not have a clinical evidence base we could find
We searched for clinical outcome trials of ultraviolet blood irradiation and did not find a body of them — largely patents and laboratory work on blood components. We offer the add-on and we price it openly on the ozone therapy page. We are not going to give it an evidence section it has not earned.
Supporting studies
-
1
The proposed mechanism is a controlled, moderate oxidative stress that activates the body's own antioxidant machinery — and the line between benefit and harm is the strength of that stress.
This review sets out the biochemical account of ozone therapy. Moderate oxidative stress activates the transcription factor Nrf2, which drives transcription of antioxidant response elements and the production of superoxide dismutase, glutathione peroxidase, glutathione-S-transferase, catalase, heme oxygenase-1 and NADPH-quinone-oxidoreductase. Severe oxidative stress instead activates NF-κB, driving COX-2, PGE2 and inflammatory cytokines. The authors draw the explicit analogy that moderate exercise is beneficial while excessive exercise is not. This is a mechanistic review, not a clinical trial, and the authors themselves call for research to establish whether these transcription factors are in fact the operative pathway.
Mechanistic review Sagai M, et al. Mechanisms of Action Involved in Ozone Therapy: Is healing induced via a mild oxidative stress? Medical Gas Research. 2011. PMID 22185664 ↗
-
2
In pooled randomised trials, intra-articular ozone gave pain control comparable to hyaluronic acid, a standard joint injection, between four and six months.
This PRISMA-compliant meta-analysis pooled randomised controlled trials comparing intra-articular hyaluronic acid against ozone in knee osteoarthritis, restricted to studies with four to six months of follow-up. Data from 424 patients were collected; 74% were women, mean age 61.1 years, mean BMI 27.8. Groups were compatible at baseline. No difference was found in visual analogue scale pain scores at follow-up (P = 0.4). The authors concluded that current level I evidence suggests ozone and hyaluronic acid achieve similar pain control over that window. Note that this is a comparison of equivalence, not of superiority.
Level I meta-analysis of randomised controlled trials Migliorini F, et al. Intra-articular injections of ozone versus hyaluronic acid for knee osteoarthritis: a level I meta-analysis. European Journal of Orthopaedic Surgery & Traumatology. 2024. PMID 39579218 ↗
-
3
Ozone outperformed placebo and performed slightly below other active injections, with adverse events occurring at a similar rate in both groups.
Five randomised clinical trials totalling 428 patients were pooled — 225 receiving ozone, 203 receiving a control injection (hyaluronic acid, dextrose or air). Mean age was 64. In the first month the pooled mean difference in VAS favoured ozone slightly (-0.23, P = 0.71); by three and six months the difference favoured the control injections (1.04 and 1.31). WOMAC scores followed the same trend. Adverse events occurred in 4% of the ozone group (6/150) and 5.4% of controls (7/129), P = 0.31. The authors concluded ozone was significantly superior to placebo and non-significantly lower than other active injections, and could be recommended as a non-surgical option durable for three to six months in mild to moderate knee osteoarthritis. Reported here in full, including the parts that do not favour ozone.
Systematic review and meta-analysis of randomised controlled trials Raeissadat SA, et al. An investigation into the efficacy of intra-articular ozone (O2-O3) injection in patients with knee osteoarthritis: a systematic review and meta-analysis. Journal of Pain Research. 2018. PMID 30498370 ↗
-
4
Across the broadest available comparison — 79 randomised trials and 8,761 patients — ozone is one credible injectable among eleven, and not the top-ranked one.
This network meta-analysis pooled 79 randomised controlled trials with 8,761 patients, comparing eleven intra-articular options: autologous conditioned serum, bone marrow aspirate concentrate, botulinum toxin, corticosteroids, hyaluronic acid, mesenchymal stem cells, ozone, saline placebo, platelet-rich plasma, plasma rich in growth factor, and stromal vascular fraction. Treatments were ranked by P-score at 1, 3, 6 and 12 months. Stromal vascular fraction achieved the highest ranking for VAS at all post-injection time points and for WOMAC at 12 months. Ozone is included in the comparison set rather than leading it. Included here specifically because it is the least flattering framing of the ozone knee data, and it belongs on the page.
Network meta-analysis of randomised controlled trials Anil U, et al. The efficacy of intra-articular injections in the treatment of knee osteoarthritis: A network meta-analysis of randomized controlled trials. The Knee. 2021. PMID 34500430 ↗
-
5
In diabetic foot ulcers, ozone alongside standard care was associated with markedly faster wound-area improvement and a substantially lower amputation rate — with no effect on complete healing or hospital stay.
Twelve randomised controlled trials were included, covering topical ozone (gas bath, ozonated oil, ozone water flushing) and systemic routes including autologous blood immunomodulation and rectal insufflation. Compared with standard control therapy for diabetic foot ulcers, ozone therapy — whether as monotherapy or combined with control treatment — accelerated improvement of wound area (standardised mean difference 66.54%, 95% CI 46.18 to 86.90, P < 0.00001) and reduced the amputation rate (risk ratio 0.36, 95% CI 0.24 to 0.54, P < 0.00001). The review explicitly found no improvement in the proportion of participants with completely healed wounds, and none in length of hospital stay. No adverse events associated with ozone treatment were reported. The authors state efficacy for other wound types remains uncertain for lack of sufficient studies.
Systematic review of randomised controlled trials Wen Q, et al. A systematic review of ozone therapy for treating chronically refractory wounds and ulcers. International Wound Journal. 2022. PMID 34612569 ↗
-
6
A second review found higher healing rates, shorter hospital stays and increased catalase activity — and warned of considerable risk of bias in the trials it pooled.
Sixty-one studies were screened and six met the eligibility criteria. Patients receiving ozone therapy displayed higher rates of healed ulcers, shorter hospital stays, increased catalase activity and reduced blood glucose levels. The authors were explicit that there was a considerable risk of bias in the studies evaluated, and concluded that although ozone therapy is a promising candidate for diabetic foot ulcers, further research with a better scientific design is required to validate the results. That caveat is theirs, and it is reproduced here rather than trimmed.
Systematic review and meta-analysis Filho MLES, et al. Ozone therapy as a treatment for diabetic foot ulcers: a systematic review and meta-analysis. Journal of Wound Care. 2024. PMID 39630555 ↗
-
7
In a randomised trial of 73 patients, 71% of those receiving ozone autohemotherapy alongside conventional care met the response threshold, versus 45% on conventional care alone.
Seventy-three participants were randomly assigned to major ozone autohemotherapy plus conventional therapy (n = 35) or conventional therapy alone (n = 38). With effectiveness pre-defined as a 50% or greater reduction in symptom score, 25 of 35 patients (71%) responded in the ozone group against 17 of 38 (45%) in the control group (P = 0.0325). Significant between-group improvements were also seen in symptom scores (P = 0.0478), tidal volume (P = 0.0374), predicted six-minute walk distance (P = 0.0032), and coagulation and inflammatory indicators. The authors describe this as a pilot trial and state that further research is warranted to validate the findings. Single-centre, 73 patients — a starting point, not a settled result.
Pilot randomised controlled trial He Y, et al. A pilot randomized controlled trial of major ozone autohemotherapy for patients with post-acute sequelae of COVID-19. International Immunopharmacology. 2024. PMID 39018686 ↗
-
8
In 40 critically ill ventilated patients, adding ozone autohemotherapy to routine care was associated with more ventilator-free days and shorter intensive-care stays.
Forty critically ill patients receiving invasive mechanical ventilation were randomised to major ozonated autohemotherapy plus routine treatment and standard critical supportive care, or to routine treatment and standard supportive care alone. The ozone group received treatment three times per week, ten times in total. Patients in the ozone group had more days breathing with no aid, and length of stay in the intensive care unit was significantly lower than in the control group. This is a 40-patient pilot trial in a single intensive care setting.
Pilot randomised controlled trial Aghamohammadi D, et al. Effects of adjuvant ozone autohemotherapy combined with routine treatment on clinical and paraclinical features of mechanically ventilated COVID-19 patients in the intensive care unit: a pilot randomized controlled trial. Medical Gas Research. 2024. PMID 37929510 ↗
-
9
For low back pain, oxygen-ozone injections placed under imaging guidance outperformed injections placed by anatomical landmarks alone.
This meta-analysis of the published literature on percutaneous oxygen-ozone injections for low back pain compared image-guided against non-image-guided injection technique. Imaging-guided procedures showed better performance than techniques relying only on anatomical landmarks, with higher therapeutic efficacy and lower age-related variability in clinical results. The finding is about procedural technique rather than about ozone itself — worth knowing precisely because it shows that how a procedure is performed is part of whether it works.
Meta-analysis Rimeika G, et al. Metanalysis on the effectiveness of low back pain treatment with oxygen-ozone mixture: Comparison between image-guided and non-image-guided injection techniques. European Journal of Radiology Open. 2021. PMID 34934778 ↗
-
10
A safety review across nine databases found no pattern of frequent serious harm when a standardised protocol is followed — and named mandatory G6PD screening as the first of those safeguards.
Following PRISMA-ScR guidance, the authors searched nine databases for adverse events associated with systemic oxygen-ozone autohaemotherapy. The literature consisted predominantly of case reports of rare but potentially serious events: haemolysis and renal failure associated with excessive ozone concentrations, hyperkalaemia in patients with complex comorbidities, ischaemic events, embolism in patients with patent foramen ovale, autonomic reactions related to rapid reinfusion rates, anaphylactic reactions linked to equipment materials, and infectious complications from protocol breaches. The authors stress that overall incidence of serious adverse events cannot be reliably quantified in the absence of prospective registries, and that frequently cited historical low estimates should be treated with considerable caution. Their conclusion is that current evidence does not identify a pattern of frequent serious unexpected harm when a standardised protocol is followed, with key safety measures being mandatory glucose-6-phosphate dehydrogenase screening, adherence to recommended ozone concentrations, controlled reinfusion rates, pre-treatment cardiovascular evaluation in selected cases, and strict aseptic technique.
Scoping review of safety and adverse events Casale R, et al. Oxygen-ozone autohaemotherapy in fibromyalgia: safety profile and adverse events. A scoping review. Clinical and Experimental Rheumatology. 2026. PMID 42328943 ↗
References
- 1Sagai M, et al. Mechanisms of Action Involved in Ozone Therapy: Is healing induced via a mild oxidative stress? Medical Gas Research. 2011. PMID 22185664 ↗
- 2Migliorini F, et al. Intra-articular injections of ozone versus hyaluronic acid for knee osteoarthritis: a level I meta-analysis. European Journal of Orthopaedic Surgery & Traumatology. 2024. PMID 39579218 ↗
- 3Raeissadat SA, et al. An investigation into the efficacy of intra-articular ozone (O2-O3) injection in patients with knee osteoarthritis: a systematic review and meta-analysis. Journal of Pain Research. 2018. PMID 30498370 ↗
- 4Anil U, et al. The efficacy of intra-articular injections in the treatment of knee osteoarthritis: A network meta-analysis of randomized controlled trials. The Knee. 2021. PMID 34500430 ↗
- 5Wen Q, et al. A systematic review of ozone therapy for treating chronically refractory wounds and ulcers. International Wound Journal. 2022. PMID 34612569 ↗
- 6Filho MLES, et al. Ozone therapy as a treatment for diabetic foot ulcers: a systematic review and meta-analysis. Journal of Wound Care. 2024. PMID 39630555 ↗
- 7He Y, et al. A pilot randomized controlled trial of major ozone autohemotherapy for patients with post-acute sequelae of COVID-19. International Immunopharmacology. 2024. PMID 39018686 ↗
- 8Aghamohammadi D, et al. Effects of adjuvant ozone autohemotherapy combined with routine treatment on clinical and paraclinical features of mechanically ventilated COVID-19 patients in the intensive care unit: a pilot randomized controlled trial. Medical Gas Research. 2024. PMID 37929510 ↗
- 9Rimeika G, et al. Metanalysis on the effectiveness of low back pain treatment with oxygen-ozone mixture: Comparison between image-guided and non-image-guided injection techniques. European Journal of Radiology Open. 2021. PMID 34934778 ↗
- 10Casale R, et al. Oxygen-ozone autohaemotherapy in fibromyalgia: safety profile and adverse events. A scoping review. Clinical and Experimental Rheumatology. 2026. PMID 42328943 ↗
On our sourcing. Every one of the 10 sources on this page was pulled directly from the live PubMed database — the US National Library of Medicine's index of biomedical literature — confirming that each PubMed ID points to the exact paper cited, with a matching author, title, journal and year. Click any reference to read the original. Where a study is preclinical, or was conducted in animals rather than people, this page says so on that study.
This page summarises published research about ozone therapy. It is educational information about that research. It is not a claim that this or any Nature & Science Medicine service diagnoses, treats, cures or prevents any disease, and it is not a substitute for a consultation with a licensed provider. Individual results vary. The FDA has not approved ozone for therapeutic use and classifies it as having no established medical application; that position and the research summarised here are both true at the same time, and we publish both.