Guide
PRP vs Cortisone for Knee Pain: Which Injection Works Better and for How Long?
For knee arthritis, a cortisone shot usually works faster but wears off sooner, while PRP (platelet-rich plasma) is slower and more expensive but may give longer relief for some people. A Cochrane review found cortisone's benefit is largest in the first few weeks and fades by about 6 months, and several recent meta-analyses report better pain scores with PRP than with cortisone at 6 to 12 months. The catch: the largest placebo-controlled PRP trial found no benefit over salt-water injections, and major guidelines disagree, so neither shot is a clear winner for everyone.
This guide walks through how each injection works, what the research actually shows, the side effects, and the questions worth asking before you choose. It is general information, not a substitute for advice from your own doctor.
What is the difference between PRP and cortisone?
Cortisone (a corticosteroid, or steroid shot) is a powerful anti-inflammatory medicine injected directly into the knee joint. It is usually mixed with a numbing medicine such as lidocaine, because the steroid itself takes a few days to kick in (AAOS OrthoInfo).
PRP is made from your own blood. A sample is drawn and spun in a centrifuge to concentrate the platelets, which carry growth factors involved in healing. According to the American Academy of Orthopaedic Surgeons (AAOS), the platelet concentration in PRP can be 5 to 10 times higher than normal blood, and the concentrated plasma is then injected into the injured area (AAOS OrthoInfo). Exactly how PRP works is still not clear.
In short, cortisone aims to calm inflammation quickly, while PRP is meant to support the joint's own healing processes over a longer period.
How fast and how long does a cortisone shot work?
Pain may drop right away because of the numbing medicine, then return for a few hours before the steroid starts working. Johns Hopkins notes it can take up to a week to notice relief, and that a cortisone shot can control inflammation and pain for several weeks or months (Johns Hopkins Medicine). AAOS also warns of a possible "flare," where pain gets worse for 1 to 2 days before the steroid takes effect.
The best summary of the research is a 2015 Cochrane review of 27 trials with 1,767 people with knee osteoarthritis (Jüni et al., Cochrane 2015). Compared with sham injections or no treatment, steroid injections reduced pain by about 1 point on a 10-point scale overall. The benefit was moderate at 1 to 2 weeks, small to moderate at 4 to 6 weeks, small at 13 weeks, and there was no evidence of an effect at 26 weeks. The reviewers rated the quality of this evidence as low, because results varied widely between trials.
Do repeated cortisone shots damage the knee?
This is one of the biggest concerns. In a 2-year randomized trial published in JAMA, 140 people with knee osteoarthritis got either a triamcinolone (steroid) injection or saline every 12 weeks (McAlindon et al., JAMA 2017). The steroid group lost more cartilage thickness on MRI (an average of 0.21 mm vs 0.10 mm) and did not have significantly less pain than the saline group.
That is why doctors limit how often you can have one. AAOS says there is no formal limit, but providers often set a practical cap of 3 to 4 shots per year. Cleveland Clinic says you will probably need to wait at least three months between rounds and that most people should not have more than three shots a year (Cleveland Clinic). Johns Hopkins adds that if one or two shots have not helped, more cortisone shots are unlikely to.
Does PRP actually work for knee arthritis?
The evidence is genuinely mixed, which is why you will hear very different answers from different doctors.
The strongest negative result is the RESTORE trial, published in JAMA in 2021 (Bennell et al., JAMA 2021). It randomized 288 people aged 50 and older with mild to moderate knee osteoarthritis to three weekly injections of leukocyte-poor PRP or saline placebo. After 12 months, pain improved by 2.1 points with PRP and 1.8 points with placebo, a difference that was not statistically significant. Cartilage volume on MRI did not differ either. The authors concluded their findings do not support PRP for knee osteoarthritis.
On the other side, AAOS's patient page says more and more literature is showing PRP is effective for low- to moderate-grade knee osteoarthritis, that some studies show results lasting up to 2 years, and that it is still unknown whether leukocyte-rich or leukocyte-poor PRP is better (AAOS OrthoInfo). It also says PRP may cause more pain at the injection site for a week or two and that it can be several weeks before patients feel a benefit.
On joint structure, a review of 12 trials we covered found PRP did not consistently protect or rebuild knee cartilage on imaging (Does PRP rebuild knee cartilage?).
Is PRP better than cortisone for knee pain?
When the two are compared head to head, PRP tends to come out ahead on longer-term pain, though the quality of the included trials varies.
- A 2025 meta-analysis of 21 randomized trials (2,406 participants) comparing PRP with placebo or corticosteroids reported that all pain scores improved more with PRP than with steroid injections (Journal of Clinical Orthopaedics and Trauma, 2025).
- A 2025 network meta-analysis of 37 randomized trials (5,089 patients) that only included studies with at least one year of follow-up ranked PRP combined with hyaluronic acid first and PRP alone second for pain and function at one year. Corticosteroids alone ranked last, behind placebo (Journal of Orthopaedic Surgery and Research, 2025).
Those results fit with the Cochrane finding that cortisone's effect fades within months. But head-to-head comparisons do not settle whether PRP beats a placebo shot, and the well-designed RESTORE trial says it may not. Comparisons with other injections are also mixed; for example, one recent trial found a dextrose-ozone-magnesium injection gave more pain relief than steroid at 6 months (read our summary).
What do medical guidelines recommend?
The 2019 American College of Rheumatology and Arthritis Foundation guideline for osteoarthritis strongly recommends intra-articular glucocorticoid (cortisone) injections for knee osteoarthritis, and strongly recommends against PRP, citing a lack of standardized preparations and high-quality evidence (Kolasinski et al., Arthritis Care & Research 2020). The same guideline strongly recommends exercise, weight loss for people who are overweight, tai chi, cane use and topical or oral NSAIDs.
AAOS's patient education is more positive about PRP for mild to moderate knee arthritis, as described above. Both organizations agree that injections are one part of treatment, not a replacement for exercise and strength work (see our guide to the best exercises for knee arthritis).
What are the side effects and risks?
Cortisone: Cleveland Clinic lists pain at the injection site, temporary swelling, bruising, skin discoloration or lightening near the injection site, and temporarily raised blood sugar as the most common side effects. People with diabetes should check their blood sugar more often for about five days afterward. Infection is rare but serious (Cleveland Clinic). Repeated shots are also linked to cartilage loss, as the JAMA trial above showed.
PRP: Because it uses your own blood, AAOS describes the risks as minimal, mainly increased pain at the injection site, with rates of infection, tissue damage and nerve injury that appear no different from cortisone shots (AAOS). A 2025 meta-analysis of 24 trials (2,751 patients) found complications in 18.7% of PRP patients vs 9.1% of patients given other injections, about one extra complication for every 11 people treated. The odds were similar to cortisone and hyaluronic acid but higher than placebo, and almost all complications were mild or moderate and went away on their own (Arthroscopy, 2025).
Does insurance cover PRP or cortisone?
Cortisone injections are a long-established, guideline-recommended treatment. PRP is a different story: AAOS advises checking with your insurer first, noting that few insurance plans, including workers' compensation plans, provide even partial reimbursement (AAOS). Ask your clinic for the full out-of-pocket price, how many injections are included, and how the PRP is prepared.
How do I decide between PRP and cortisone?
Questions worth discussing with your doctor:
- Do you need fast relief? For a painful flare, a trip or an event, cortisone works within days.
- How many cortisone shots have you already had? If you are near the 3 to 4 per year limit, or shots have stopped working, other options make sense.
- How severe is your arthritis? Most PRP research focuses on mild to moderate (Kellgren-Lawrence grade 2 to 3) knee osteoarthritis.
- Do you have diabetes? Cortisone can raise blood sugar for several days.
- Can you afford PRP if insurance does not pay, knowing a high-quality placebo-controlled trial found no benefit?
- Are you also doing exercise therapy? Every major guideline puts exercise and weight management first.
See a doctor promptly if, after any knee injection, you develop fever, rapidly worsening swelling, redness or intense pain, which can signal infection.
Frequently asked questions
- How long does a cortisone shot in the knee last?
- Usually a few weeks to a few months. A Cochrane review found the pain benefit was greatest at 1 to 2 weeks, small by 13 weeks, and no longer detectable at 26 weeks.
- How long does it take PRP to work in the knee?
- AAOS says it may be several weeks before patients feel a benefit, and pain at the injection site may increase for the first week or two.
- Can you get PRP after a cortisone shot?
- The sources reviewed for this guide do not set a standard waiting period. Ask the doctor doing the PRP how long they want you to wait after your last steroid injection.
- How many cortisone shots can you get in your knee?
- There is no formal limit, but providers often cap them at 3 to 4 per year (AAOS), and Cleveland Clinic advises waiting at least three months between rounds.
- Is PRP safer than cortisone?
- PRP avoids steroid side effects such as raised blood sugar, and AAOS says its risks are minimal. But a 2025 meta-analysis found PRP had more mild complications than placebo, with odds similar to cortisone and hyaluronic acid.
- Does PRP regrow knee cartilage?
- There is no reliable evidence that it does. The RESTORE trial found no difference in cartilage volume vs placebo at 12 months, and a review of 12 trials found inconsistent structural effects.
Sudden severe pain, swelling, or can't bear weight? Seek medical care promptly. In an emergency call 911.