Knee Injections Compared: Cortisone, Hyaluronic Acid and PRP

A 58-year-old hiker told me last month that she had watched the same clip four times: a professional footballer, sleeve rolled up, describing how a shot of his own blood “rebuilt” his knee. By the end of the week she had asked her orthopedic team about it, and so, judging by search data, had a great many other people. As of September 2026, interest in knee injections is running well above its usual autumn peak, driven by athlete testimonials about platelet-rich plasma, a steady drumbeat of clinic marketing for “regenerative” treatments, and a guideline landscape that has quietly shifted against one long-standing option.

The frustrating part is that the three injections most often offered for a worn knee have very different track records, and the loudest claims attach to the one with the thinnest evidence. This piece lays out what each injection actually does inside the joint, how long relief tends to last, who should steer clear, and how confident the science really is.

What are the three types of knee injections?

When people search for knee injections, they almost always mean treatments for osteoarthritis, the wear-related condition in which the smooth cartilage cushioning the joint thins and the surrounding tissues become inflamed and painful. Three injections dominate that conversation, and each enters the knee the same way, through a needle into the joint space itself. Clinicians call this an intra-articular injection, meaning the medication is placed inside the joint capsule rather than into muscle or skin.

Cortisone is the oldest and most widely used. It is a corticosteroid, a synthetic version of the anti-inflammatory hormone cortisol that your adrenal glands make naturally. In the knee it dampens the inflammatory chemistry that drives swelling and pain. Hyaluronic acid, often called a gel injection, is a thick, slippery molecule that occurs naturally in healthy joint fluid; the injected version is meant to restore some of that lubrication. Platelet-rich plasma, usually shortened to PRP, is made from a sample of the patient’s own blood, spun in a centrifuge to concentrate platelets, the tiny cell fragments that carry growth factors involved in healing.

Those three sit on very different regulatory footings. Corticosteroid injections are approved medicines with decades of use. Hyaluronic acid products are regulated in the United States as medical devices approved for knee osteoarthritis. PRP is prepared with cleared equipment, but PRP itself has not been approved as a treatment for knee osteoarthritis, and in England the national guideline says it should not be offered outside research. You may also hear about newer options, including stem-cell style injections; those are investigational, are not approved for routine use, and are not something to seek out from a commercial clinic on your own.

What changed recently with knee injections

The biggest shift is in official advice about gel injections. The health service guidance for England, updated in late 2022 and still current in 2026, states that hyaluronic acid injections should not be offered to manage osteoarthritis, because pooled trial data did not show a benefit large enough to justify routine use. Corticosteroid injections kept their place, but only as a short-term option when other measures have not given enough relief. That is a notable change for a treatment that was offered freely for years.

Cortisone’s own story has also been revised by evidence. A randomized trial published in 2017 gave people with knee osteoarthritis repeated steroid injections or saline every few months for two years. Pain was no better in the steroid group at the end, and imaging showed slightly more cartilage loss. Mainstream patient guidance now carries that caution plainly: repeated cortisone may weaken cartilage and nearby tissue, which is why clinicians space the shots out and limit how many they give.

For PRP, the key date is 2021, when the largest placebo-controlled trial to that point compared PRP with saline injections in nearly 300 people. At twelve months there was no meaningful difference in pain or in cartilage volume on MRI. Smaller studies since have pointed in both directions, which is exactly why PRP remains labelled as uncertain rather than proven.

Why, then, is PRP trending upward? The 2026 surge traces to social media rather than to a new trial: a cluster of athlete endorsements and clinic promotions presenting PRP as a way to “regrow” cartilage. Nothing in the peer-reviewed literature supports that phrasing. What has not changed is the strong, consistent evidence for exercise and weight management as the foundation of knee osteoarthritis care, which every major guideline still places ahead of any needle.

How a cortisone shot in the knee works, and how long it lasts

Think of an inflamed knee as a room where the smoke alarm will not switch off. Cortisone does not fix the wiring; it silences the alarm for a while. Once inside the joint, the corticosteroid slows production of the chemical messengers that recruit immune cells, dilate blood vessels and sensitize nerve endings. Swelling eases, warmth fades and pain drops, sometimes dramatically.

Timing matters for expectations. Many people notice less pain within a day or two, though some feel a brief flare in the first 24 to 48 hours as the crystals of medication irritate the joint lining before they dissolve. Relief then typically holds for a few weeks. Randomized trials show the advantage over placebo is clearest at around four to six weeks and usually fades by three months. Some people get longer, some shorter; a knee with obvious fluid and heat tends to respond better than one that aches without visible inflammation.

Because the effect is temporary and because repeated exposure carries risks to cartilage, bone and skin, cortisone is best understood as a bridge rather than a destination. The classic use is to calm a painful flare enough that a person can start, or restart, a strengthening program, or to get through an important stretch, say a wedding or a long-planned trip. Your clinician will decide how often it can be repeated based on your joint, your other conditions and your response the first time; the schedule is theirs to set, not something to negotiate up.

People with diabetes should know that a corticosteroid injection can raise blood glucose for several days. It is worth mentioning that condition before the appointment so monitoring can be planned.

Gel injections for knee pain: what hyaluronic acid does

Healthy knee fluid, called synovial fluid, is thick enough to coat a spoon, and hyaluronic acid is the ingredient that gives it that quality. In osteoarthritis the fluid becomes thinner and less protective. The idea behind viscosupplementation, the technical name for gel injections, is intuitive: top up the lubricant and the joint should glide more smoothly.

The mechanism turns out to be less straightforward than the brochure suggests. Injected hyaluronic acid is cleared from the joint within days, far sooner than any reported benefit, so if it helps, it probably does so by nudging the joint lining to make more of its own hyaluronic acid or by dampening inflammatory signaling, not by acting as a long-lived cushion. Products differ in molecular weight and in whether they are given as a single injection or as a short series over consecutive weeks; the course is product-specific and determined by the prescribing clinician.

What people actually experience varies widely. Some report a gradual easing of pain over several weeks that lasts for months. Others feel nothing. Trials capture both, which is why the pooled result is a small average benefit that many experts consider below the threshold of what a patient would notice. Temporary soreness and swelling at the injection site are the most common side effects; a more pronounced inflammatory reaction, sometimes called a pseudoseptic flare, occurs occasionally and needs assessment to rule out infection.

Gel injections remain available and used in many countries, and some orthopedic societies still see a role for them in selected patients who cannot take anti-inflammatory tablets or who want to delay surgery. The honest framing is that they are safe, modest and unpredictable, and that the person best placed to judge whether they are worth trying is the clinician who has examined your knee.

PRP knee injection: what is actually in the syringe

A PRP knee injection starts with a blood draw from your arm, usually a few tubes’ worth. The sample goes into a centrifuge that spins it fast enough to separate red cells from the lighter plasma, and within that plasma the platelets are concentrated several times above their normal level. That platelet-rich layer is drawn into a syringe and injected into the knee, often within the same hour.

Platelets matter because they release growth factors, proteins that signal cells to multiply, migrate and lay down new tissue. In a cut finger, that cascade is part of normal healing. The hope is that delivering a concentrated dose into an arthritic joint will calm inflammation and support cartilage cells. In laboratory dishes, growth factors do stimulate cartilage cells. Whether that translates into a thicker cartilage layer inside a living human knee is the question the trials have been asking, and so far the answer from the most rigorous studies is no.

One reason PRP research is hard to summarize is that PRP is not one thing. Preparations differ in platelet concentration, in whether white blood cells are included, in whether the mixture is activated before injection, and in how many injections make up a course. Two clinics offering “PRP” may be injecting quite different products, which makes it difficult to compare studies or to know which protocol, if any, a positive trial actually supports.

Because PRP uses your own blood, allergic reactions are essentially unheard of, and infection risk is similar to any joint injection. The main downsides are a few days of post-injection soreness, the time involved, and the risk of paying for hope. If you are considering it, ask the clinician what preparation they use, what the published evidence shows for that preparation, and what realistic outcome they expect.

What the evidence actually says, graded

Evidence in medicine comes in tiers. At the top sit randomized controlled trials, studies in which people are assigned by chance to a treatment or a placebo so that the only systematic difference is the treatment itself. Below them are observational studies, which follow people who chose a treatment and compare them with those who did not; useful, but prone to bias. At the bottom is expert opinion. Here is where each knee injection lands.

Cortisone: moderate-certainty evidence of short-term benefit. Dozens of randomized trials, pooled in systematic reviews, show that a single corticosteroid injection reduces knee osteoarthritis pain more than placebo for a few weeks. The effect size is meaningful early and shrinks to nothing by about three months. For long-term repeated use, the best trial evidence shows no sustained pain benefit and a small signal of cartilage loss.

Hyaluronic acid: low-certainty evidence of small benefit. There are many trials, but the larger and better-blinded ones tend to find less benefit than the smaller ones, a pattern that usually points to bias inflating the early results. The most recent large pooled analyses conclude that the average pain reduction versus placebo is small and possibly not clinically important. Serious adverse events are rare.

PRP: low to very-low certainty, with conflicting results. Several small trials and meta-analyses report benefit, often compared with hyaluronic acid rather than placebo. The largest saline-controlled trial found no difference at twelve months on pain or cartilage. Heterogeneity in preparations and short follow-up limit confidence in either direction.

One finding cuts across all three: the placebo response to any injection into the knee is large. Saline injections alone produce measurable pain relief in trials, probably through a mix of expectation, the ritual of the procedure, and the physical effect of fluid in the joint. Any honest comparison has to subtract that effect first.

Knee injections compared: side-by-side table

Laid out together, the differences are easier to see. The table summarizes what mainstream guidance and trial data say about each option for knee osteoarthritis; it is a starting point for a conversation with your clinician, not a substitute for one.

Feature	Cortisone	Hyaluronic acid (gel)	PRP
What it is	Synthetic anti-inflammatory steroid	Lubricating molecule found in joint fluid	Concentrated platelets from your own blood
Proposed mechanism	Suppresses inflammatory signaling	May prompt joint lining to restore lubrication; mild anti-inflammatory effect	Growth factors may reduce inflammation and support cartilage cells
Typical onset	Hours to days	Several weeks	Weeks
Typical duration of any relief	Weeks, rarely beyond three months	Months in responders; many feel nothing	Uncertain; reported months in some studies
Strength of evidence	Moderate for short-term relief; repeated use not supported	Low; small average effect of doubtful importance	Low to very low; largest placebo trial negative
Regulatory status (US)	Approved medicine	Approved medical device for knee osteoarthritis	Not approved as a treatment; prepared with cleared devices
Guidance in England	Consider for short-term relief when other measures insufficient	Not recommended	Research use only
Main cautions	Blood glucose rise, cartilage effects with repetition, infection	Post-injection flare, allergy to product components	Variable preparations, soreness, unproven claims

Two patterns stand out. First, confidence in the evidence runs in the opposite direction to the volume of online enthusiasm: the most promoted option is the least proven. Second, none of the three has been shown to change the course of osteoarthritis. They are pain-management tools, with different speeds and different reliability, and the structural condition of the joint carries on regardless.

What is the most effective injection for knee pain?

If “effective” means the greatest chance of noticeable relief in the next month, cortisone wins, and the evidence behind that is reasonably solid. If it means the longest-lasting relief, the honest answer is that no injection has a dependable long-term record, and the trial data for hyaluronic acid and PRP are too inconsistent to crown either one.

The more useful question is which injection fits which knee. A joint that is hot, swollen and has fluid you can feel is in an inflammatory flare, and that is the scenario in which a corticosteroid is most likely to help quickly. A knee that grinds and aches with activity but is not particularly swollen is a weaker candidate for cortisone, and this is where some clinicians still consider a gel injection for people who cannot take anti-inflammatory tablets or want to postpone surgery. PRP is mostly discussed for younger or more active people with mild to moderate changes on X-ray, partly because the small trials that showed benefit tended to recruit that group; that is a hypothesis, not an established indication.

Severity matters too. Across all three injections, people with advanced osteoarthritis, where the cartilage has largely worn through and bone is rubbing on bone, tend to respond least. At that stage the realistic conversation is often about surgery rather than about which injection to try next.

There is also a strong case for thinking about what happens after the injection. The strongest predictor of how a person with knee osteoarthritis feels a year from now is not which needle they chose but whether they built strength in the muscles around the joint and managed their weight. An injection that buys a window to do that work can be worth having. An injection that substitutes for that work usually leads back to the same waiting room.

How long do knee injections last?

Duration is the question people most want answered and the one trials answer least neatly, because averages hide enormous individual variation. With that caveat, the published patterns look like this.

Cortisone acts fastest and fades fastest. Benefit over placebo is clear at two to six weeks and generally gone by three months. Some people report a single shot settling a flare that then stays quiet for much longer, but in those cases the flare was probably going to settle anyway; cortisone shortened the storm rather than preventing the next one.

Hyaluronic acid is slower to start, with any effect usually building over four to eight weeks. In people who respond, relief has been reported to last around six months in some trials, which is the basis for repeating a course after that interval. Because a substantial share of recipients notice no benefit at all, a quoted average can mislead.

PRP timelines come mostly from smaller studies, which report improvements lasting six to twelve months in some participants. Set against that is the largest placebo-controlled trial, in which the PRP and saline groups were indistinguishable at twelve months. If PRP does provide durable relief for some people, the field has not yet identified who they are in advance.

Several factors shorten relief for any injection: advanced joint damage, obesity, ongoing heavy loading of the knee, and inflammation driven by a condition other than osteoarthritis. The tissue the needle enters is also the tissue that bears your body weight with every step, so relief tends to last longer when the weeks after the injection are used to reduce load and improve strength rather than to resume everything that hurt before.

Who should not get knee gel injections, and who should avoid the others

Every intra-articular injection shares one absolute rule: no needle goes into a knee if there is any suspicion of infection in the joint or in the skin over it. Introducing bacteria into a joint can cause septic arthritis, a rapidly destructive infection that is a surgical emergency. Broken skin, cellulitis or an unexplained hot, red joint all postpone the appointment.

For gel injections specifically, people with a known allergy to hyaluronic acid products should not receive them. Some older formulations were derived from rooster combs, and manufacturers historically advised caution in people with egg or poultry allergy; many current products are made by bacterial fermentation, but the prescribing clinician will check the specific product. A knee with a large effusion, meaning a build-up of fluid inside the joint, is usually drained first, since injecting gel into an already distended joint can increase pain. Because gel injections are not recommended in several national guidelines, your clinician may simply not offer them.

Cortisone carries its own list. Poorly controlled diabetes is a relative contraindication because glucose can rise for days. A recent or planned joint replacement is another: many surgeons avoid a steroid injection in the months before surgery because of a possible link to post-operative infection. People on blood thinners can usually be injected but need the clinician to know. Repeated injections into the same knee are limited, and anyone who has had several already should expect a conversation about alternatives.

PRP is unsuitable for people with platelet disorders, certain blood cancers or active infection, and it is generally not advised during anticoagulant therapy without specialist input. Anti-inflammatory tablets are often paused around the injection because they may blunt the platelet response, but stopping any prescribed medicine is a decision for the prescriber, not something to do on your own.

Side effects and risks of knee injections, honestly ranked

Most knee injections are uneventful. The needle stings, the joint feels full for a day, and life goes on. The risks that remain are worth ranking by how often they happen and how much they matter.

The most common complaint across all three options is a post-injection flare: increased pain, warmth and swelling that begins within hours and settles within a couple of days. It affects a minority of patients and is usually managed with rest and ice. With cortisone it can be mistaken for the medication failing; in fact the steroid effect typically arrives after the flare passes.

Cortisone has a longer tail of effects specific to steroids. Blood glucose rises in people with diabetes. Facial flushing can occur for a day or two. At the injection site, the skin may thin or lighten and a small dimple of fat loss can appear, particularly if medication leaks outside the joint. Repeated injections are associated in trial data with faster cartilage loss and, in some observational studies, with accelerated progression of osteoarthritis, which is why guidelines limit frequency and total number.

Hyaluronic acid’s notable risk is the pseudoseptic reaction, an intense inflammatory flare that mimics infection. It is uncommon but needs prompt assessment, because the two cannot be told apart without examining the joint and sometimes testing the fluid.

PRP has the shortest list of drug-type effects because the material is your own blood. The main risks are soreness that can last longer than after cortisone, bruising at the blood-draw site, and the possibility of being promised a result that the evidence does not support.

The rare, serious risk shared by all three is joint infection, estimated in large series at well under one in a thousand injections when sterile technique is used. It is uncommon enough not to be a reason to avoid injection, and serious enough that the warning signs, covered below, deserve a place in your memory.

Common myths about knee injections, corrected

The clips driving this year’s search spike contain a handful of claims that recur so often they deserve a direct answer.

“PRP regrows cartilage.” The most rigorous evidence available, including the largest placebo-controlled trial with MRI follow-up, found no difference in cartilage volume between PRP and saline. Laboratory studies show growth factors can stimulate cartilage cells in a dish; a living joint under load behaves differently. At present, cartilage regrowth from PRP is a hypothesis, not a demonstrated effect.

“Cortisone is dangerous and should never be used.” The evidence supports short-term use for flares and argues against repeated long-term use. That is a nuance, not a ban. A single well-placed injection to break a painful cycle remains a reasonable, guideline-endorsed option.

“Gel injections replace your joint fluid.” Injected hyaluronic acid is cleared from the knee within days. If it helps, it does so indirectly, not by sitting in the joint as a cushion for months.

“If one injection helped, a stronger one will help more.” There is no dose-escalation ladder for knee injections. Higher steroid doses do not produce longer relief and increase tissue risk. Repeating PRP or gel courses is sometimes done, but the decision rests on your response and your clinician’s assessment.

“Natural means safe, so PRP has no risks.” Infection risk is the same as any joint injection, and the preparation step adds variability. “From your own body” does not mean “proven.”

“Injections let you skip exercise.” Every major guideline places exercise and weight management ahead of injections, because they are the interventions with the most consistent long-term benefit. Injections work best as a way to make that work possible, not as a replacement for it.

Injections are one tool: what else moves the needle for knee osteoarthritis

It is easy to lose sight of the fact that the interventions with the strongest and most durable evidence for knee osteoarthritis do not involve a syringe. Structured exercise, particularly strengthening of the quadriceps and hip muscles, reduces pain and improves function in trial after trial, and the effect does not wear off at three months. Supervised programs delivered by a physical therapist tend to outperform leaflets, largely because people stick with them.

Weight matters more than most people assume, because the knee experiences several times body weight with each step. Trial data show that losing a modest proportion of body weight produces measurable reductions in pain and improvements in walking, and the gains scale with the amount lost. This is not about appearance; it is about load through a joint that has lost some of its shock absorption.

Simpler measures help too. A cane used in the opposite hand unloads a painful knee noticeably. Supportive footwear, a knee brace for some patterns of arthritis, and heat or ice during flares all have a place. Topical anti-inflammatory gels have reasonable evidence and fewer systemic effects than tablets; oral anti-inflammatories help but carry stomach, kidney and cardiovascular cautions that make them a short-term tool for many people. Which, if any, suits you is a decision for your clinician, particularly if you take other medicines.

When pain remains limiting despite all of this, joint replacement is the intervention with the most reliable long-term results for advanced osteoarthritis, and most people report substantial improvement in pain and mobility afterward. Injections sit in the middle of this ladder. They are most valuable when they open a window to climb the lower rungs, and least valuable when they become a loop of repeat appointments that delays a more definitive conversation.

When to see a doctor

Two kinds of situations call for medical attention around knee injections: problems after a procedure, and knee symptoms that should be assessed before anyone considers injecting at all.

After any knee injection, contact the clinic or seek urgent care if you notice signs that could indicate joint infection. Treat these as red flags rather than things to watch for a few days:

• Pain that worsens sharply after the first 48 hours instead of easing.
• A knee that becomes hot, red and tense, especially with fever, chills or feeling generally unwell.
• Inability to bear weight or to bend the knee because of pain.
• Spreading redness, pus or a wound that will not close at the injection site.
• In people with diabetes, blood glucose readings that stay far above your usual range after a cortisone injection.

Septic arthritis can damage a joint within days, so the threshold for a same-day assessment should be low. A pseudoseptic flare after gel injection looks similar and also needs examination, because the two cannot be distinguished at home.

Before injections come up at all, certain knee symptoms deserve a proper evaluation rather than a trial of any treatment: a knee that locks or gives way, sudden swelling after a twist or fall, a fever with joint pain, pain that wakes you at night and does not vary with activity, or unexplained weight loss alongside joint symptoms. These point away from straightforward osteoarthritis and toward problems that need imaging, blood tests or a specialist opinion.

Finally, every decision described in this article, which injection, whether to repeat it, whether to pause a medicine around it, or whether to move toward surgery, belongs with the clinician who has examined your knee and knows your history. Bring questions, bring the clips that prompted them if you like, and expect an honest conversation about what the evidence can and cannot promise.

Frequently asked questions

What are the three types of knee injections?

The three main knee injections for osteoarthritis are cortisone, hyaluronic acid and platelet-rich plasma. Cortisone is a corticosteroid that calms inflammation. Hyaluronic acid, often called a gel injection, mimics a lubricating molecule in joint fluid. PRP is made from your own blood, spun to concentrate platelets. All three are injected directly into the joint space, and they differ sharply in how well their benefits are proven.

What is the most effective injection for knee pain?

For short-term relief, cortisone has the strongest evidence, with benefit over placebo lasting a few weeks. No injection has reliably demonstrated long-term relief. Hyaluronic acid shows small, inconsistent effects, and the largest placebo-controlled PRP trial found no advantage at one year. Which option, if any, fits your knee depends on how inflamed it is, how advanced the arthritis is, and your clinician’s assessment.

Who should not get knee gel injections?

Gel injections should be avoided by anyone with an infection in or around the knee, a known allergy to hyaluronic acid products, or a large build-up of fluid that has not been drained. Some older products carried cautions for people with egg or poultry allergies. Because several national guidelines no longer recommend hyaluronic acid for osteoarthritis, your clinician may suggest other options instead.

How many times can you get gel injections in the knee?

The number of gel injections in a course is set by the specific product: some are designed as a single injection, others as a short series over consecutive weeks. Repeat courses are sometimes considered after several months if the first helped. There is no fixed lifetime limit in the evidence, but the decision to repeat rests with the prescribing clinician based on your response and joint condition.

How long does a cortisone shot in the knee last?

Relief from a cortisone shot in the knee usually begins within a day or two and lasts a few weeks. Randomized trials show the clearest benefit over placebo at four to six weeks, with little difference by three months. Some people get longer relief, particularly when the injection calms an inflammatory flare. Clinicians space repeat injections apart because repeated steroid exposure is linked to cartilage loss.

Is a PRP knee injection better than cortisone?

Not according to the strongest evidence. Some small trials report PRP outperforming cortisone or hyaluronic acid over several months, but these studies vary widely in how PRP was prepared. The largest placebo-controlled PRP trial found no benefit over saline at twelve months. Cortisone has more consistent proof of short-term relief; PRP’s long-term promise remains unproven. Your clinician can explain what the evidence means for your specific knee.

How long do knee injections last compared with each other?

Cortisone acts fastest and fades fastest, typically within weeks. Hyaluronic acid starts more slowly and, in people who respond, may last around six months, though many notice no effect. PRP studies report six to twelve months in some participants, but the largest placebo-controlled trial found no difference from saline at one year. Advanced arthritis, obesity and heavy joint loading all shorten relief from any injection.

Do knee injections damage the joint over time?

Repeated cortisone injections are the main concern. A two-year randomized trial found slightly more cartilage loss with regular steroid injections than with saline, and observational studies have linked frequent steroid use to faster osteoarthritis progression. A single injection for a flare has not been shown to cause harm. Hyaluronic acid and PRP have not shown structural damage, though neither has shown structural benefit either.

What are the red flags after a knee injection?

Seek urgent assessment if pain worsens sharply after 48 hours, the knee becomes hot, red and tense, you develop fever or chills, you cannot bear weight, or there is pus or spreading redness at the injection site. These can signal joint infection, which needs same-day treatment. People with diabetes should also report blood glucose that stays well above their usual range after a cortisone injection.

Can knee injections replace surgery for osteoarthritis?

No injection has been shown to change the underlying course of osteoarthritis, so they cannot replace joint replacement for advanced disease. They can delay the decision by managing pain for a period, which some people value. For severe, bone-on-bone arthritis, injections tend to work least well, and joint replacement has the most reliable long-term results. The timing of that step is a decision for you and your surgeon.

References

• Osteoarthritis: Treatment, NHS
• Osteoarthritis: MedlinePlus
• Osteoarthritis: NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.