If you have a painful knee, shoulder, elbow, or heel, you will eventually be offered an injection. The two most common options sit at opposite ends of almost every axis: cost, speed, regulatory status, and the quality of the evidence behind them.

This guide is information, not medical advice. It can't tell you which shot your joint needs — only what the published research does and doesn't support.

The short version

Corticosteroid ("cortisone") injections are an FDA-approved drug, usually covered by insurance, and reliably reduce pain for a few weeks to a few months. The benefit fades, and repeated use carries real trade-offs for cartilage and tendon tissue.

Platelet-rich plasma (PRP) is your own blood, spun down to concentrate platelets and the growth factors they carry, then injected. It does little for the first few weeks. In some conditions — tennis elbow is the best example — it appears to outperform steroid at six months and beyond. In others, high-quality placebo-controlled trials have found nothing. It is almost never covered by insurance.

The honest framing is not "which is better" but "what time horizon are you buying, and what are you willing to pay and risk for it."

One is an approved drug. The other isn't.

Injectable corticosteroids such as triamcinolone and methylprednisolone are FDA-approved drugs with labeled indications, known dosing, and a documented adverse-event profile.

PRP is not an FDA-approved treatment for any orthopedic condition. The FDA clears the centrifuges and kits used to prepare PRP as devices, which is a different and much lower bar — it certifies that the device concentrates platelets as described, not that injecting the result helps your knee. The agency has published consumer guidance warning that many marketed regenerative products have not been shown to be safe or effective.

That does not mean PRP is illegal or fringe. A physician can lawfully use it, and it's offered in mainstream orthopedic and sports medicine practices. It does mean there is no approved label defining the right dose, the right preparation, or the right number of injections — which is a large part of why the trial results are so inconsistent. (Our separate guide on how PRP is regulated goes further into this.)

The timing difference is the whole story

Corticosteroid is an anti-inflammatory. It suppresses the pain signal quickly, typically within days, and in most studies that advantage is largest in the first one to six weeks and has narrowed or disappeared by three to six months.

PRP is proposed to work by modulating the local healing environment. Whatever it does, it does slowly. Trials that show benefit generally show it at 3, 6, or 12 months — not at two weeks. If you need to function next Tuesday, PRP is the wrong tool.

This crossover pattern shows up most clearly in tendon studies. A randomized trial of PRP versus corticosteroid for tennis elbow found steroid ahead early, with PRP ahead at one year (Peerbooms et al., American Journal of Sports Medicine, 2010).

Tendons: where steroid may actively work against you

This is the sharpest distinction in the literature. A systematic review of injection treatments for tendinopathy found corticosteroid superior in the short term but associated with worse outcomes and higher recurrence at intermediate and long-term follow-up compared with other approaches or no injection (Coombes et al., The Lancet, 2010).

A later randomized trial in tennis elbow made the point bluntly: patients who received a corticosteroid injection had lower complete recovery rates at one year and more recurrences than those who received placebo injection (Coombes et al., JAMA, 2013).

That finding has changed practice. Many clinicians now avoid steroid in chronic tendinopathy, and injection directly into a load-bearing tendon — the Achilles especially — is generally avoided because of concern about tendon weakening and rupture.

PRP is not an automatic winner by default, though. In a sham-controlled randomized trial of chronic midportion Achilles tendinopathy, PRP produced no significant benefit over sham injection at six months (Kearney et al., JAMA, 2021). The tendon evidence is condition-specific: reasonably encouraging for lateral epicondylitis and plantar fasciitis, unimpressive for Achilles, thin for most others.

And in essentially every tendinopathy trial, the comparison group is also doing progressive loading exercise. Structured rehab is the treatment with the most consistent support. An injection of either kind is an adjunct.

Joints: smaller, blurrier differences

Inside a joint, the trade-offs are different. Steroid has a genuine but brief effect on osteoarthritis pain, and Cochrane reviewers have repeatedly rated the certainty of that evidence as low, with benefit hard to demonstrate past a few weeks.

The repeat-use question got a direct answer in a two-year randomized trial: patients receiving intra-articular triamcinolone every 12 weeks had significantly greater cartilage volume loss than those receiving saline, with no pain advantage at two years (McAlindon et al., JAMA, 2017). One trial doesn't settle the matter, but it is the reason many clinicians now think carefully before scheduling steroid on a standing quarterly basis.

PRP in joints is genuinely contested. Pooled analyses have suggested benefit over saline or hyaluronic acid at 6–12 months, but the single largest rigorous placebo-controlled trial, RESTORE, found no difference from saline injection in pain or cartilage volume at 12 months (Bennell et al., JAMA, 2021). We cover that debate, dosing schedules, and what a realistic course looks like in PRP for knee osteoarthritis: what to expect from a course — rather than repeat it here.

Outside the knee, the joint evidence thins quickly. For hip, ankle, and small-joint osteoarthritis, PRP trials are fewer, smaller, and less consistent. Treat any claim of equivalent evidence across joints with suspicion.

How often can you repeat either one?

Steroid. The widely used convention is no more than three to four injections per joint per year, spaced at least three months apart. This is clinical custom informed by the cartilage and soft-tissue concerns above, not a number derived from a dose-finding trial. Other known effects include a post-injection pain flare, skin lightening or fat atrophy at the injection site, and a transient blood sugar rise that matters if you have diabetes. If joint replacement is on your horizon, tell your surgeon about recent injections — timing relative to surgery is a recognized consideration.

PRP. There is no established limit, because there is no approved label and no long-term safety database. Common practice is one to three injections spaced two to four weeks apart, sometimes with a booster at six to twelve months. Since the material is autologous, serious adverse events are rare in the published trials, but "few reported harms in short studies" is not the same as demonstrated long-term safety. The practical ceiling is usually financial.

A detail worth asking about: PRP preparations differ enormously in platelet concentration and in whether white blood cells are included. Leukocyte-rich and leukocyte-poor PRP are not interchangeable, and researchers have hypothesized they suit different tissues. Clinics rarely disclose which they use. Ask.

Cost and coverage

A corticosteroid injection is typically billed as a routine office procedure and covered by most insurance; out of pocket it commonly runs from under a hundred dollars to a few hundred, depending on imaging guidance and facility fees.

PRP is usually self-pay. Single injections commonly range from roughly $500 to $2,000, and a multi-injection course can reach $1,500 to $4,000 or more. Prices vary with imaging guidance, preparation system, and region far more than with any measurable difference in the product. Insurers generally classify musculoskeletal PRP as investigational.

A reasonable way to think about it

If you need relief quickly, have an inflammatory flare, and the target is a joint rather than a tendon, steroid is the cheaper, better-characterized option — used sparingly.

If you have chronic tendinopathy that has already failed a real course of loading exercise, PRP is a defensible next step, with the understanding that you are paying out of pocket for a modest and condition-dependent chance of improvement over several months.

If a clinic quotes you a success rate, or offers either injection without first asking what rehab you've done, that's a signal about the clinic rather than about the treatment.