In short
A PRP injection for knee pain concentrates platelets from your own blood and places them in the joint. Serious complications are uncommon, but the best placebo-controlled trials have not shown a clear benefit, and guidelines treat it as unproven rather than standard care.
A PRP injection for knee pain uses your own blood. It is drawn from a vein, spun in a centrifuge to concentrate the platelets, and the concentrate is injected into the knee joint. Because it comes from you, there is no risk of rejection and no donor to worry about.
The honest summary of the evidence is that PRP appears safe and that its benefit is uncertain. NICE, reviewing the published evidence, concluded that platelet-rich plasma injections for knee osteoarthritis raise no major safety concerns but that the evidence on how well they work is limited in quality (NICE).
That is an unsatisfying answer when your knee hurts. This article sets out what the trials found, why they disagree with one another, what is genuinely unknown, and what to ask before agreeing to a course, so that whichever way you decide, you decide it knowing where the evidence stands.
What is actually in a PRP injection
Blood is a mixture of red cells, white cells, platelets and plasma. Spinning it separates those layers by weight. The fraction that is drawn off contains platelets at several times their normal concentration, suspended in plasma.
Platelets are best known for clotting, but they also carry granules full of growth factors, the signalling proteins released at the site of an injury to start repair. Concentrating them and placing them inside the joint is an attempt to borrow that repair signal.
It is worth being clear about what PRP is not. It is not a stem cell treatment, it contains no cells taken from anyone else, and it is not a steroid. It also is not the same product from one clinic to the next, which turns out to matter a great deal.
The idea behind injecting platelets into a knee
Osteoarthritis is not simply a joint surface wearing away. The joint lining becomes inflamed, the bone underneath remodels, and the whole joint behaves like a struggling organ rather than a worn bearing. Growth factors that damp inflammation and influence the cells that maintain cartilage are therefore a reasonable thing to try.
Reasonable is not the same as proven. Laboratory work shows platelet-derived growth factors affect cartilage cells in a dish, and animal work has shown effects on joint tissue. Whether that translates into less pain in a human knee a year later is exactly the question the clinical trials were built to answer.
One claim to be wary of is cartilage regrowth. When a trial measured cartilage volume on MRI scans directly, it found no advantage over a placebo injection. Nothing available today reliably regrows lost cartilage in an arthritic knee, and a treatment described that way is being oversold.
What a PRP injection for knee pain involves on the day
The procedure. Blood is taken from your arm as for any blood test. It is spun, usually for several minutes, while you wait. The concentrate is then injected into the knee, ideally with ultrasound guidance so that the needle is demonstrably inside the joint space rather than in the fat pad beside it. The injection itself takes a minute or two.
Afterwards. Soreness and a feeling of fullness in the knee for two or three days is common and expected, more so than after a steroid injection. Most people walk out and drive home. Strenuous activity is usually deferred for a short period, and the knee is then used normally.
The course. Protocols differ. Some give a single injection, others a series spaced over some weeks. There is no settled standard, which is itself a sign of how unsettled the evidence is. If you already take anti-inflammatory tablets, ask before the appointment whether the clinic wants them paused, and speak to your rheumatologist before changing anything you have been prescribed.
How joint injections in general are done, and when they are worth doing, is covered on the page about joint aspiration and injection.
What the trials show
The most informative trials are those comparing PRP against a placebo injection, because a needle entering a painful knee produces a real and sizeable improvement on its own.
In the RESTORE trial, adults with mild to moderate knee osteoarthritis received either PRP or a saline placebo injection. The authors reported that PRP did not result in a significant difference in symptoms or joint structure at twelve months, and concluded that the findings do not support the use of PRP for the management of knee osteoarthritis (RESTORE trial, via PMC).
Set against that, a body of smaller trials compares PRP with hyaluronic acid or with a steroid injection rather than with placebo, and a number of them report better pain and function scores with PRP, particularly in earlier-stage disease. A recent review summarising this literature notes those findings alongside the caveat that heterogeneity in PRP formulations, injection regimens and follow-up periods limits direct comparison, and that high-quality placebo-controlled trials show inconsistent long-term benefit (narrative review, via PMC).
Both of those statements are true at once. PRP often performs better than another injection whose own benefit is modest, and struggles to beat salt water in the best-designed trials.
The placebo response in a painful knee is not a trick of the mind, either. Fluid entering the joint, the attention of an unhurried appointment, the expectation of improvement and the natural tendency of a flare to settle all push the score in the same direction. Any treatment given by injection starts with that tailwind, which is precisely why the comparison has to be against another injection rather than against no treatment.
Where the uncertainty really lies
The disagreement between studies is not random. It follows from how the studies were built.
- Comparator. Beating hyaluronic acid is a lower bar than beating a placebo injection, and the two designs answer different questions.
- Severity. Trials that recruit early-stage arthritis tend to report better results than those including advanced joint damage.
- Follow-up. Differences that look convincing at three months often shrink by twelve, and most trials stop long before the disease does.
- Outcome chosen. Pain scores, function scores and imaging do not always move together, and a trial can be positive on one and flat on the others.
None of this means the treatment is a fraud. It means the honest position is uncertainty, and that anyone telling you the question is settled, in either direction, has gone beyond the evidence.
This is also why two careful doctors can read the same literature on a PRP injection for knee pain and give you different advice, both defensibly. One weighs the placebo-controlled trials most heavily and concludes the benefit is unproven. The other weighs the comparative trials in early disease and concludes it is worth trying in selected people before surgery is discussed. Neither is misleading you.
Why no two preparations are alike
PRP is a category, not a product. Preparations differ in how concentrated the platelets are, whether white cells are included or deliberately excluded, whether the platelets are activated before injection, and how many spins the blood goes through.
Those differences are biologically meaningful: white cells bring inflammatory signals of their own, which some researchers argue is unhelpful inside an arthritic joint. When trials using different preparations are pooled together, the result is an average of treatments that are not really the same treatment.
The practical consequence for you is simple. Ask what system the clinic uses and whether the preparation is leukocyte-rich or leukocyte-poor. A clinic that knows exactly what it is injecting is on firmer ground than one that does not.
What the guidelines say
NICE reviewed the procedure and advised that it should be used only with special arrangements for clinical governance, consent and audit or research, and that clinicians should ensure people understand the uncertainties about how well it works before agreeing to it (NICE).
The broader NICE guideline on osteoarthritis keeps the emphasis elsewhere. Therapeutic exercise and weight management are named as the core treatments; a topical anti-inflammatory is offered first for knee osteoarthritis; intra-articular hyaluronan injections are not recommended; and steroid injections are described as giving short-term relief only, of the order of two to ten weeks (NICE).
PRP sits outside that standard pathway rather than replacing any part of it. If it is used, it is used alongside the things that have stronger evidence, not instead of them.
Nothing in either document forbids a PRP injection for knee pain. What they ask is that the person receiving it understands that the benefit is not established, that the decision is recorded properly, and that the rest of the treatment plan is not quietly dropped in the meantime.
Who is least likely to benefit
Advanced structural damage. When the joint space has largely gone, the knee is deformed, or walking distance is severely limited, an injection of any kind is unlikely to change much. That is the point at which a surgical opinion is the more useful conversation.
A knee that is not osteoarthritic at all. Persistent knee swelling with more than thirty minutes of morning stiffness, pain in several joints, or a knee that flares and settles in episodes may be inflammatory arthritis, gout or something else again. Injecting platelets into an inflammatory joint treats the wrong problem. Telling rheumatoid arthritis and osteoarthritis apart sets out the pattern that separates them, and the wider picture of arthritis and joint pain is worth reading before committing to any procedure.
A knee with a mechanical block. True locking, or giving way that throws you off balance, points towards a torn meniscus or a loose fragment, which is an orthopaedic question.
Risks, and the things that genuinely go wrong
Serious complications are uncommon, which is the part of the evidence that is actually reassuring. The realistic list is short.
- A flare of pain and swelling for a few days after the injection, which settles.
- Bruising at the blood draw, or feeling faint while blood is taken.
- Infection in the joint. Rare, but the reason sterile technique matters.
Treat this as an emergency. A knee that becomes hot, very painful and swollen in the days after any injection, particularly with fever or shaking chills, needs same-day assessment and usually aspiration of the joint fluid. Septic arthritis destroys a joint quickly, and it is not something to watch over a weekend. The same applies to any hot swollen joint with fever whether or not you have had an injection, as when to see a rheumatologist explains.
Questions worth asking before you agree
- What preparation is used, and is it leukocyte-rich or leukocyte-poor?
- How many injections are planned, over what period, and will they be ultrasound-guided?
- What will we measure, and at what point will we agree it has not worked?
- Has anything other than osteoarthritis been ruled out as the cause of this knee pain?
- What is the plan for exercise, strengthening and weight alongside the injections?
That last question is the one that changes outcomes most reliably. Quadriceps strengthening and weight management are the interventions with the strongest evidence in osteoarthritis of the knee, and they keep working long after any injection has worn off. Exercise and weight in arthritis covers how to start when the joint already hurts, and what helps in osteoarthritis of the knee, in what order sets out the full sequence.
If you decide to go ahead with a PRP injection for knee pain, go ahead with your eyes open, a clear review date and a strengthening programme running alongside. If you decide against it, you have not missed a proven treatment. Either way, bring the knee to a proper assessment first, so that what is being injected is at least the right diagnosis.
Common questions
Does PRP regrow cartilage in the knee?
There is no reliable evidence that it does. A trial that measured cartilage volume on MRI found no difference between PRP and a placebo injection after twelve months. Any benefit people get is more likely to come from reduced inflammation and pain than from new cartilage being built.
Is PRP better than a steroid injection for knee arthritis?
Some smaller trials suggest better scores with PRP over several months, while steroid injections work faster and fade sooner. The comparison is unsettled because trials use different preparations and follow people for different lengths of time. Neither changes the underlying arthritis.
How many PRP injections are needed for a knee?
There is no agreed number. Some protocols use a single injection, others a short series spaced over weeks. The absence of a standard reflects genuine uncertainty rather than a secret formula. Ask the clinic what it plans, and agree in advance when you will judge whether it helped.
Is a PRP injection safe?
Serious problems are uncommon. The usual effects are a few days of soreness and swelling, bruising where blood was taken, and occasionally feeling faint. Joint infection is rare but serious. A hot, very painful, swollen knee with fever after any injection needs same-day medical assessment.
Sources
- NICE - Platelet-rich plasma injections for knee osteoarthritis www.nice.org.uk
- NICE - Osteoarthritis in over 16s: diagnosis and management www.nice.org.uk
- JAMA (PMC) - The RESTORE randomised trial of PRP versus placebo pmc.ncbi.nlm.nih.gov
- Journal of Clinical Medicine (PMC) - PRP for knee osteoarthritis review pmc.ncbi.nlm.nih.gov
This article is general information and does not replace a consultation.