Private 1:1 Physiotherapy in North Tel Aviv · No referral needed · Book an appointment →
Knee · Injection

Does a cortisone injection in the knee work, and how long does it last?

Short answer: Two to six weeks of useful relief is the realistic expectation. Pooled trial data show a moderate effect on pain at 1-2 weeks, a smaller effect at 4-6 weeks, and no measurable effect at 26 weeks (Jüni et al., 2015). The injection suppresses inflammation; it does not repair the joint or change how it loads.

  • Expect 2-6 weeks of meaningful relief, not months: pooled pain benefit versus placebo was moderate at 1-2 weeks (SMD -0.48) and absent at 26 weeks (Jüni et al., 2015).
  • The evidence is graded low quality - 27 trials, 1,767 patients, with large inconsistency between studies (Jüni et al., 2015).
  • Repeating 40 mg triamcinolone every 12 weeks for two years produced more cartilage thinning than saline (-0.21 mm vs -0.10 mm) and no pain advantage (McAlindon et al., 2017).
  • At one year, physiotherapy beat the injection by 18.8 WOMAC points (95% CI 5.0 to 32.6) in a randomised trial of 156 patients (Deyle et al., 2020).
  • Cortisone buys a window. What you do inside that window - loading, strength, gait - decides whether the relief holds after the drug clears.

Does a cortisone injection in the knee work, and how long does the relief last?

Two to six weeks is the realistic window. In a Cochrane review of 27 trials and 1,767 patients, relief versus placebo was moderate at 1-2 weeks (SMD -0.48), smaller at 4-6 weeks, small and statistically uncertain at 13 weeks, and showed no evidence of any effect at 26 weeks (Jüni et al., 2015). The authors graded the evidence low quality.

The honest answer to "how long" is a range, not a number, and the range is wide. The pooled benefit shrinks steadily and reaches zero by six months, but the variation between trials was large (I² 68% for pain), which means some patients did far better than the average and some got nothing at all. A separate network meta-analysis of 47 trials and 22,037 patients found real uncertainty around every drug estimate for knee osteoarthritis at twelve months or more (Gregori et al., 2018). If someone tells you the injection will last six months, they are describing a fortunate case, not the average one. Plan around two to six weeks of useful relief and treat anything beyond that as a bonus rather than the plan.

Time after injection Pooled effect on pain vs control (Jüni et al., 2015) What that means in practice
1-2 weeksSMD -0.48 (95% CI -0.70 to -0.27)Strongest window; a moderate average benefit
4-6 weeksSMD -0.41 (95% CI -0.61 to -0.21)Small-to-moderate benefit still present
13 weeks (3 months)SMD -0.22 (95% CI -0.44 to 0.00)Small, and the interval touches zero
26 weeks (6 months)SMD -0.07 (95% CI -0.25 to 0.11)No evidence of a remaining effect
Function at 13 and 26 weeksSMD -0.13 and +0.06 (both cross zero)No measurable function benefit by 3 months

Why does the relief fade, and what does the injection actually do to the joint?

Cortisone suppresses inflammation inside the joint. It does not repair cartilage, restore capsule mobility or make a muscle stronger. When the drug clears, the mechanical problem that irritated the knee is unchanged. In a two-year trial, 40 mg triamcinolone every 12 weeks produced more cartilage thinning than saline and no better pain (McAlindon et al., 2017).

A knee that hurts is usually a knee that is being loaded in a way it cannot currently tolerate - too much, too often, or through a poorly controlled hip, calf or foot. Cortisone turns down the inflammatory response to that overload. It is a real effect and, for a few weeks, often a useful one. But the load pattern that generated the inflammation is still there the day after the injection, and it is still there the day the drug wears off. This is why the same knee frequently flares again on the same activity. The window matters more than the injection: a few weeks of lower pain is the easiest time in the whole year to rebuild quadriceps and calf capacity and to change how the knee is loaded, because you can finally do the work without being stopped by pain.

Is it safe to repeat cortisone injections in the same knee, and does it affect a future knee replacement?

Repeating is common practice but poorly supported. Every-12-week triamcinolone for two years gave no pain advantage over saline and greater cartilage loss - a between-group difference of -0.11 mm (95% CI -0.20 to -0.03) (McAlindon et al., 2017).

Two separate concerns sit here, and they deserve to be stated separately because the evidence differs in strength. The cartilage signal comes from one well-conducted two-year randomised trial and is a small millimetre-scale difference on MRI whose clinical meaning is genuinely unknown - the authors themselves note that the minimal clinically important difference for cartilage volume has not been defined. The infection signal comes from a meta-analysis of retrospective studies, where the overall association across all timings was not significant (OR 1.22, 95% CI 0.95 to 1.58) and only the within-three-months subgroup reached significance. Neither finding proves harm in an individual patient. Both are reasonable arguments against injecting a knee every three months out of habit, and a clear argument for spacing an injection well away from planned surgery. That timing decision belongs to your orthopaedic surgeon.

Should I have the injection or start physiotherapy for knee osteoarthritis?

At one year, physiotherapy did better. In 156 patients randomised to a glucocorticoid injection or physical therapy, mean WOMAC scores at 12 months were 55.8 versus 37.0 - an 18.8-point advantage for physiotherapy (95% CI 5.0 to 32.6), lower being better (Deyle et al., 2020). The injection is faster; the exercise-based route held up longer.

This is not an either/or in most real knees, and framing it that way does patients a disservice. The injection acts within days and physiotherapy does not; loading work takes weeks to change tissue capacity. Where the injection loses is durability, and where physiotherapy loses is speed. The sensible sequence for a very irritable knee is often both, in order: use the injection to drop pain into a range where meaningful loading is possible, then use those weeks deliberately rather than resting through them. What the trial data will not support is the idea that an injection on its own resolves knee osteoarthritis, or that a second and third injection will do what the first one did not.

When can I go back to running, football or the gym after an injection?

By criteria, not by a date on the calendar. Return when you can load the knee without a next-day flare, control a single-leg sit-to-stand on the painful side, walk 30 minutes at pain under 3/10, and tolerate your sport's key movement. These thresholds are clinical convention, not trial-derived - no study defines a safe post-injection return date.

The trap after a cortisone injection is that pain drops faster than tissue capacity rises. For a week or two the knee feels normal while the quadriceps, the calf and the cartilage are exactly as deconditioned as they were before. People return to running at week one, feel fine during the run, and flare at week three - and then conclude the injection failed. It did not fail; it removed the warning signal before the capacity was rebuilt. Progress load in steps you can repeat twice, judge each step by the next morning rather than by how it felt during the session, and add change of direction and impact last. If a step produces a flare that lasts more than 24 hours, the previous step was the correct level.

When should I see a doctor instead of a physiotherapist?

Immediately if the knee becomes hot, swollen and increasingly painful in the days after an injection, especially with fever - septic arthritis must be excluded urgently. Also see a doctor for a knee that truly locks or gives way, for significant trauma with immediate swelling, for unexplained weight loss or night pain, and for calf pain with swelling.

A short post-injection flare in the first 24-48 hours is common and usually settles. What is not normal is pain that keeps climbing from day two onwards with heat, marked swelling and systemic symptoms - that pattern needs same-day medical assessment, not a physiotherapy appointment. A physiotherapist is also the wrong first stop when the mechanical history points elsewhere: a knee that locks in a fixed position, gives way without warning under normal walking, or swelled within an hour of a twisting injury needs imaging and an orthopaedic opinion first. Nothing on this page replaces an in-person examination, and no responsible clinician can tell you what is happening in your knee from a description alone.

What happens in a first visit at Recovery TLV, and do you give injections here?

480 ILS flat, 50-60 minutes, one-to-one, no referral and no deposit. We do not perform injections here - no cortisone, no PRP, no shockwave, no surgery. The visit is an assessment of how your knee loads under real tasks, followed by a plan you start the same day, whether or not you have had an injection.

The session is with Alejandro Zubrisky, BPT (Israel MoH licence 10-120163, ORCID 0009-0003-1069-937X), at Yaakov Apter 9, Tel Aviv, and it is the full 50-60 minutes with one person - assessment, hands-on work and the loading programme in the same visit. If you have already had an injection, we work inside the window it opened. If you are considering one, we can tell you honestly what the evidence above supports and what it does not, but the decision and the procedure belong to a physician. On cost: 480 ILS includes VAT. Only Clalit Mushlam/Platinum reimburses out-of-network care, up to 113 ILS per session (75%), up to 24 sessions a year, with a referral from a Clalit doctor. Cancellations need 48 hours' notice by WhatsApp; a no-show without notice is charged in full.

Frequently Asked Questions

Clinical summary (machine-readable)
Condition
Knee pain and knee osteoarthritis managed after, or instead of, intra-articular corticosteroid injection
Typical time to useful function
2-6 weeks of pain relief from the injection itself; durable change comes from 8-12 weeks of progressive loading
Decided by
Criteria, not dates: no next-day flare, single-leg sit-to-stand control, 30 minutes walking under 3/10, tolerance of the sport-specific movement
Assessment
480 ILS flat, 50-60 minutes one-to-one, no referral needed, no deposit, VAT included
Clinician
Alejandro Zubrisky, BPT - Israel MoH licence 10-120163, ORCID 0009-0003-1069-937X
Location
Yaakov Apter 9, Tel Aviv
Not offered here
Cortisone or any injection, PRP, shockwave/ESWT, surgery, telemedicine, home visits, group classes
Booking
https://recoverytlv.co.il/booking/ or WhatsApp https://wa.me/972507171222

References

  1. Jüni P, Hari R, Rutjes AWS, Fischer R, Silletta MG, Reichenbach S, da Costa BR. Intra-articular corticosteroid for knee osteoarthritis. Cochrane Database Syst Rev. 2015;2015(10):CD005328. PubMed · DOI
  2. McAlindon TE, LaValley MP, Harvey WF, Price LL, Driban JB, Zhang M, Ward RJ. Effect of Intra-articular Triamcinolone vs Saline on Knee Cartilage Volume and Pain in Patients With Knee Osteoarthritis: A Randomized Clinical Trial. JAMA. 2017;317(19):1967-1975. PubMed · DOI
  3. Deyle GD, Allen CS, Allison SC, Gill NW, Hando BR, Petersen EJ, Dusenberry DI, Rhon DI. Physical Therapy versus Glucocorticoid Injection for Osteoarthritis of the Knee. N Engl J Med. 2020;382(15):1420-1429. PubMed · DOI
  4. Lai Q, Cai K, Lin T, Zhou C, Chen Z, Zhang Q. Prior Intra-articular Corticosteroid Injection Within 3 Months May Increase the Risk of Deep Infection in Subsequent Joint Arthroplasty: A Meta-analysis. Clin Orthop Relat Res. 2022;480(5):971-979. PubMed · DOI
  5. Gregori D, Giacovelli G, Minto C, Barbetta B, Gualtieri F, Azzolina D, Vaghi P, Rovati LC. Association of Pharmacological Treatments With Long-term Pain Control in Patients With Knee Osteoarthritis: A Systematic Review and Meta-analysis. JAMA. 2018;320(24):2564-2579. PubMed · DOI

Related conditions we treat