PRP vs. Cortisone Injections: Mechanism, Duration, and Patient Outcomes

Cortisone is an anti-inflammatory whose relief wears off; it is strongly recommended for knee and hip arthritis, though not on a fixed schedule. PRP is recommended against for knee and hip arthritis, and a large knee trial found it no better than a saline placebo. Tendon problems give a different answer: there, PRP has beaten cortisone in the mid term, though in the one tendon condition where it has also been tested widely against placebo, the difference was small. The answer depends on the condition and on what PRP is compared with, so this page takes them one at a time.

How each injection works

Cortisone is a corticosteroid. OrthoInfo calls corticosteroids “powerful anti-inflammatory agents that can be injected into the joint”, and says the injections “can provide pain relief and reduce inflammation; however, the effects do not last indefinitely” (OrthoInfo). Mayo Clinic says the relief “can last up to several months” (Mayo Clinic).

Platelet-rich plasma (PRP) is made from your own blood. It is drawn, spun in a centrifuge, and the platelet layer is concentrated; OrthoInfo puts the platelet concentration at “5 to 10 times greater (or richer) than usual” (OrthoInfo). The idea is that growth factors released by platelets support healing. That is a rationale, not proof; what matters for you is whether patients do better, which is what the trials below test.

How do they compare for knee arthritis?

Sources for the table: ACR; Mayo Clinic; OrthoInfo on knee arthritis; OrthoInfo on PRP; <a href="https://www.wikijournalclub.org/wiki/EffectofIntra-articularTriamcinolonevsSalineonKneeCartilageVolumeandPaininPatientsWithKneeOsteoarthritis”>cortisone trial summary; RESTORE trial.

The table is about knee arthritis only. Tendon problems are a separate question with separate trials, covered further down, and the two should not be read into each other.

What does the evidence show for knee arthritis?

Cortisone. In a two-year trial of 140 people, knees injected with triamcinolone every 12 weeks lost slightly more cartilage than knees injected with saline (0.21 mm against 0.10 mm), and pain was no better (<a href="https://www.wikijournalclub.org/wiki/EffectofIntra-articularTriamcinolonevsSalineonKneeCartilageVolumeandPaininPatientsWithKneeOsteoarthritis”>trial summary). The US National Institute of Arthritis and Musculoskeletal and Skin Diseases summarized it as showing “no long-term benefit” for shots given every three months (NIAMS). That is evidence against a fixed schedule of injections, not against an occasional shot for a flare. OrthoInfo says your doctor “may recommend limiting the number of injections to three or four per year, per joint” (OrthoInfo, above).

PRP. The RESTORE trial gave 288 people with mild to moderate knee arthritis three weekly injections of either a leukocyte-poor PRP or saline (RESTORE trial abstract). At 12 months pain had improved in both groups with no difference between them, and there were no differences on MRI. The lead researcher said the results “do not support the use of platelet-rich plasma for people with mild to moderate knee osteoarthritis” (University of Melbourne, on the JAMA trial).

Why “PRP beat cortisone” is the wrong test. A clinic may cite trials in which PRP did better than cortisone months later. Cortisone’s relief is expected to fade by then, and repeated cortisone did no better than saline for pain in the two-year trial. Beating a treatment that has worn off does not show that PRP works; comparing it with a placebo does, and in RESTORE it did not. Not every source reads the knee evidence the same way. The American Academy of Orthopaedic Surgeons’ 2021 overview concluded that PRP “may have more benefit than placebo” for knee arthritis symptoms for up to a year, while calling the effect on the joint itself unclear (<a href="https://www.aaos.org/globalassets/quality-and-practice-resources/biologics/technology-overviewprp-for-knee-oa.pdf”>AAOS technology overview, 2021); the rheumatology guideline recommends against PRP. Why major sources disagree, including differences in method and timing, is set out in the overview of PRP and the conditions it has been tried for. The rest of the knee choices sit alongside PRP in the options for a painful knee and the trials behind each.

Do tendon problems give a different answer?

They do, and the reason is instructive. A 2025 review pooled 27 randomized trials with 1,779 patients comparing PRP with cortisone for rotator cuff problems, tennis elbow, plantar fasciitis and tenosynovitis. It concluded that PRP’s “mid-term efficacy is superior to that of corticosteroids. However, the long-term efficacy remains to be further clinically verified”. The authors also note: “The majority of studies omitted details on blinding” (BMC Musculoskeletal Disorders, 2025), so patients and assessors may have known who got which injection.

The review compared PRP with cortisone, not with a placebo. Tennis elbow shows why that matters: it is one of the conditions in the review, and when a Cochrane review pooled placebo-controlled trials in tennis elbow, PRP or blood injections improved pain by 0.16 points on a 10-point scale at three months, and the reviewers concluded the data “do not support the use of autologous blood or PRP injection” (<a href="https://www.cochrane.org/CD010951/MUSKELautologous-blood-or-prp-injection-lateral-elbow-pain”>Cochrane, 2021). The same injection can look better than cortisone and about the same as placebo, because cortisone’s benefit fades. In the Achilles, one PRP injection did no better than a sham needle at six months (Kearney, JAMA 2021).

So for a tendon problem, the useful questions are which tendon, and compared with what. A result in the shoulder does not carry over to the heel, and a result against cortisone does not tell you how PRP compares with no injection. Someone with tennis elbow who has had two cortisone shots, each helping for a while, is deciding between a third shot and PRP. The cortisone comparison speaks to whether PRP might hold up better than that third shot; the placebo-controlled trials speak to whether PRP does anything on its own, and in tennis elbow they found almost nothing. Both belong in the decision.

In OrthoInfo’s advice for Achilles tendinitis, exercise comes before any injection: “Physical therapy is very helpful”, and surgery “should be considered only if the pain does not improve after 6 months of nonsurgical treatment” (OrthoInfo). Cortisone near a tendon has its own risk, listed below.

What are the risks of each?

Cortisone. Mayo Clinic lists cartilage damage, a short-term increase in blood sugar, tendon weakening or rupture, death of nearby bone, joint infection, nerve damage, thinning of nearby bone, and thinning of skin and soft tissue around the injection site (Mayo Clinic, above). OrthoInfo adds that pain and swelling may flare right after the shot, and warns: “With frequent repeated injections, or injections over an extended period of time, joint damage can actually increase rather than decrease” (OrthoInfo, above). If you have diabetes, tell the doctor before the injection and ask how to watch your blood sugar afterward.

PRP. OrthoInfo says of PRP, “The risks associated with it are minimal.” There may be “increased pain at the injection site”, and the rate of infection, tissue damage and nerve injury “appears to be no different from that associated with cortisone injections” (OrthoInfo, above). Being made from your own blood does not make it risk-free; an injection into or near a joint can let germs in (Mayo Clinic on joint infections). The signs of infection that need a prompt call, and what to do about your usual medicines, are covered under the weeks after an injection and the signs that need a call.

Can you have both?

Some clinics give cortisone for a bad flare and PRP later. Protocols differ on how long to wait between them, and none of the sources on this page sets a waiting time. Ask for the written protocol and the reason behind the gap, and if two different doctors are involved, tell both and let them agree.

What does each cost you?

OrthoInfo says few insurance plans reimburse even part of the cost of PRP, so expect to pay for it yourself unless your insurer confirms otherwise. The price that matters is the full course, including any repeat injections, and your insurer’s answer is worth having in writing before you pay for either shot, using the steps for asking your insurer before an injection. The questions for a first appointment include the cost questions to raise in person.

What is the strongest case for PRP?

Across rotator cuff problems, tennis elbow, plantar fasciitis and tenosynovitis, the 2025 review found PRP’s mid-term results better than cortisone’s, and for the knee a mainstream orthopaedic overview reads the evidence more favorably than the rheumatology guideline does. If you have done a proper exercise program, still have pain and would rather not keep having steroid shots, that evidence is worth raising with your doctor. The limits are that the tendon comparison was with cortisone rather than placebo, blinding was unclear in most of those trials, and the largest placebo-controlled trials in the knee, the Achilles and the elbow found a small difference or none.

What if neither works?

For knee arthritis, the RESTORE researchers point to “exercise and, if they are above a healthy weight range, weight loss” (University of Melbourne, above), and the guideline strongly recommends exercise (ACR, above). If an injection of either kind did not help, the more useful next step is to re-check the diagnosis than to move to a more expensive injection. A clinic that proposes the pricier option first is showing one of the signs of a clinic selling more than the evidence supports.

This page is general information, not medical advice.