PRP works for some conditions and not others, and the trials and reviews we read show which. The evidence is most favorable in knee arthritis, where an 18-trial meta-analysis found PRP better than placebo, though one large saline-controlled trial did not. Tennis elbow is contested: no better than placebo in a 2021 Cochrane review, better than cortisone at three and six months on low-quality evidence. It is mixed for plantar fasciitis, the rotator cuff, hamstring strains, outer hip pain and hip arthritis, and weak for Achilles tendinopathy and ankle arthritis.
You have read that PRP is a breakthrough, and that it is a placebo. Here are the trials, failed ones included, so you know where your injury sits before you spend money or months.
What the trials found, and where they failed
Knee arthritis: what do the trials show?
An 18-trial meta-analysis found PRP better than placebo for pain and function, the pain benefit clinically meaningful at three and six months. A 2021 meta-analysis of 18 level 1 trials found WOMAC scores improved 44.7% with PRP versus 12.6% with hyaluronic acid; another favored PRP over cortisone at three, six and nine months. Against it, a 2021 network meta-analysis ranked steroids and hyaluronic acid above PRP, and the 2021 RESTORE trial of 288 people found no difference from saline at 12 months. More on PRP for knee osteoarthritis.
Tennis elbow: why is it contested?
Placebo and cortisone tell different stories. A 2021 Cochrane review, pooling 10 placebo-controlled trials, found blood or PRP injections probably give little or no clinically important benefit over placebo, at three months and beyond. On low-quality evidence, a 2023 meta-analysis found cortisone better at one month and PRP better at three and six. More on PRP for tennis elbow.
Why are five conditions mixed?
Wins fade, or trials disagree. For plantar fasciitis, a 2024 meta-analysis of 21 trials found PRP relieved pain more than cortisone or placebo, without a consistent edge in function. For rotator cuff tendinopathy and partial tears, a 99-patient trial found PRP ahead of cortisone at three months and level at twelve. For acute hamstring strains, six pooled trials put return to play 8.6 days sooner with PRP, with high heterogeneity; a 90-athlete trial found no benefit over intensive rehabilitation. For outer hip pain, a nine-study review found PRP often ahead of cortisone, but a 79-patient trial found it no better than placebo. For hip arthritis, PRP beat hyaluronic acid for pain in a 2026 meta-analysis of seven studies, while a 2023 meta-analysis found no clear pain benefit.
Where has PRP failed?
Two JAMA trials in 2021 were negative: a single PRP injection was no better than a sham needle for chronic Achilles tendinopathy in 240 patients, and two ultrasound-guided injections were no better than saline for ankle arthritis in 100. Two recent meta-analyses, from 2025 and 2026, found no benefit over placebo in Achilles tendinopathy. A 2023 meta-analysis did find a pooled ankle pain benefit; we give the placebo-controlled trial more weight.
Why do the verdicts differ?
Partly the tissue, partly the product. Across knee trials that reported dose, the 28 of 31 PRP arms that worked at six months averaged 5.5 billion platelets; the three that did not averaged 2.3 billion. White cells are less settled: one knee meta-analysis favored leukocyte-poor PRP, a double-blind knee trial found no difference, and tendinopathy trials favored leukocyte-rich PRP. The rest is in what actually determines PRP results and high-dose versus standard PRP.
The verdict, condition by condition
Favorable: meta-analyses favor PRP over placebo. Contested: placebo and cortisone comparisons disagree. Mixed: partial wins. Weak: good trials found no benefit.
- Knee arthritis: favorable, debated, ahead of placebo and other injections in meta-analyses, behind hyaluronic acid and steroids in one network analysis, level with saline in one trial
- Tennis elbow, chronic: contested, no better than placebo in a Cochrane review, ahead of cortisone at three and six months on low-quality evidence
- Plantar fasciitis: mixed, better pain relief, no consistent gain in function
- Rotator cuff tendinopathy or partial tear: mixed, ahead of cortisone at three months, level at twelve
- Hamstring strain: mixed, faster return in pooled trials, no benefit in a 90-athlete trial
- Outer hip pain: mixed, often ahead of cortisone, no better than placebo in one trial
- Hip arthritis: mixed, ahead of hyaluronic acid in a 2026 meta-analysis, no clear pain benefit in another
- Achilles tendinopathy, ankle arthritis: weak, no clear benefit in placebo-controlled trials
Before you book, ask
- Which trials cover my condition, and were any negative?
- What platelet dose and white cell content will I get?
- When and how will we judge whether it worked?
If your condition is in the weak column
Not a dead end. PRP should not be the first thing sold to you; we will say so and show what fits.
Frequently asked questions
Does PRP actually work?
For some conditions yes, and for others the controlled trials say no. The most favorable evidence is in knee arthritis, where an 18-trial meta-analysis of 1,995 patients found PRP better than placebo injections for pain and function. In chronic tennis elbow, a 2021 Cochrane review found little or no benefit over placebo, though PRP beat cortisone at three and six months in a 2023 meta-analysis. Placebo- or sham-controlled trials were negative in Achilles tendinopathy and ankle arthritis. So the useful question is whether PRP works for your tissue, with the preparation you are being offered.
Is PRP just a placebo?
Not in every condition, but in some it has not beaten one. In knee arthritis, an 18-trial meta-analysis found PRP better than placebo injections at every follow-up from one to twelve months. In chronic tennis elbow, a 2021 Cochrane review, pooling 10 placebo-controlled trials, found little or no clinically important benefit. In Achilles tendinopathy and ankle arthritis, controlled trials found no difference, and a 79-patient trial found the same for outer hip pain. Placebo effects are real, which is exactly why the controlled trials are the ones we read.
What conditions does PRP work best for?
Knee arthritis has the most favorable evidence, ahead of placebo, hyaluronic acid and cortisone in meta-analyses, though one network analysis ranked hyaluronic acid and steroids above it and one large trial found no difference from saline. Tennis elbow is contested: no better than placebo in a Cochrane review, better than cortisone at three and six months on low-quality evidence. Plantar fasciitis, rotator cuff tendinopathy, hamstring strains, outer hip pain and hip arthritis sit in the middle. Midportion Achilles tendinopathy and ankle arthritis have negative placebo- or sham-controlled trials. We use that map at the evaluation, then examine you and review your imaging to confirm what is actually driving the pain. How we graded: for each condition we read the first 40 systematic reviews and meta-analyses (or all, where fewer) that a PubMed search returned, ranked by relevance, and the trials named here. It is a careful reading, not a formal systematic review.
Why did the big PRP trials in JAMA come back negative?
They tested specific preparations in specific tissues, and PRP did not beat the control. RESTORE gave three weekly injections of a commercial leukocyte-poor PRP for knee arthritis. The Achilles trial gave one injection into the tendon. The ankle trial gave two ultrasound-guided injections. Their published abstracts do not report platelet dose, so we will not claim dose explains them. What they do show is that PRP is not a general-purpose fix, and a clinic should tell you where it has failed.
Does PRP work for knee arthritis?
For knee arthritis the pooled evidence leans favorable but is contested. An 18-trial meta-analysis of 1,995 patients found PRP better than placebo injections, with the pain benefit clinically meaningful at three and six months, and clinically meaningful at twelve with high-platelet PRP. Meta-analyses also favor PRP over hyaluronic acid and over cortisone, the latter at three, six and nine months, while a 2021 network meta-analysis ranked steroids and hyaluronic acid above PRP, though in its direct comparisons for pain hyaluronic acid did better than single PRP and steroids did worse. Against saline, the 288-person RESTORE trial found no difference at 12 months, and an 11-trial meta-analysis found a pain benefit at three and six months only. Across trials that reported dose, the PRP arms that worked averaged more than twice the platelets of the three that did not.
How long does it take to know if PRP worked?
Most patients notice meaningful change somewhere in the three to six week window, and results continue to accrue over roughly six to twelve weeks. One 230-patient tennis elbow trial found PRP no better than needling at 12 weeks and better at 24, among the 119 followed that long, though a Cochrane review found no clinically important benefit over placebo at six or twelve months. If nothing at all has changed by six weeks, that is worth reviewing rather than waiting out, because the diagnosis, placement or dose may need a second look.
Does the type of PRP matter?
The evidence is mixed. In knee arthritis, leukocyte-poor PRP produced better IKDC knee scores than leukocyte-rich PRP in a 2021 meta-analysis, but a 132-patient double-blind trial in 2024 found no difference between the two, while in tendinopathy the strongest effects came from leukocyte-rich preparations in a 2017 meta-analysis of 18 trials. Dose tracked with results too: in knee trials that reported platelet dose, the 28 of 31 PRP arms that worked at six months averaged 5.5 billion platelets, against 2.3 billion for the three that did not. Dynamic PRP+ delivers over 10 billion platelets, concentrated 12 to 20 times baseline.
Is PRP worth it if insurance does not cover it?
It can be, when your condition sits in the favorable column and the preparation matches the trials. PRP is out of pocket, and Cherry financing is available. HSA and FSA dollars typically apply to physician-delivered procedures when they are prescribed for a specific medical indication. The honest comparison is not PRP against zero. It is a result that lasts against more months of pain, missed training and repeat visits. We give a quote after a thorough evaluation. We price the plan for optimal outcomes, not line items off a menu. If PRP is a weak fit, we say that first.
Find out which column your injury is in
Book an evaluation with Aneesh Garg, DO, CAQ. He examines the tissue and tells you what the evidence says. A strong fit can buy back a season; a poor fit costs months. Every Dynamic PRP+ injection is performed by him under live ultrasound.
Book an Evaluation
1790 30th Street Suite 270 · Boulder, CO 80301 · 303-997-1733