The short answer: most sources will tell you three to four injections a year in the same area, spaced at least six to twelve weeks apart. That ceiling is real and it is reasonable. But it is a practical guardrail built around cumulative steroid exposure — it is not a biological law, and it is not how the decision should actually be made.
The number of injections you need is based on your clinical need, and above all on how you responded to the first one. There is no particular magic in a series of three.
Where the “three a year” number actually comes from
Corticosteroids are potent anti-inflammatory drugs. Injected near an inflamed nerve root or into an irritated joint, they can reduce swelling and interrupt a pain cycle that oral medication cannot reach. They are genuinely useful. They also have a dose-dependent cost: repeated steroid exposure at the same site is associated over time with weakening of local tendon, ligament and cartilage tissue, and repeated systemic absorption affects blood sugar, blood pressure and the body’s own cortisol production.
The three-or-four-per-year convention exists to cap that cumulative exposure. It is a sensible ceiling. What it is not is a schedule — and the difference matters, because a ceiling tells you when to stop, while a schedule tells you to keep going. Being told at your first visit that you will be having three injections is a booking convention, not a clinical finding. Nobody can know at injection one whether you will need injection two.
The single biggest factor is how you responded to the first injection
If the first injection gave you nothing at all, the odds that a second or third will help are slim. That is not a failure and it is not a dead end — it is diagnostic information. An injection placed correctly at a target that does not respond is telling you the target was probably wrong. The right next step is to re-examine the diagnosis, not to repeat the same injection and hope.
If the first injection helped — even partially, even briefly — that changes the calculation entirely. Partial relief means the target was right and the question becomes one of degree and durability. In that situation a repeat injection is reasonable, and there is evidence that repeating relatively early in people who responded can produce greater and longer-lasting relief than waiting a long interval.
So the two patients are not the same patient. One should probably not be injected again. The other might reasonably be injected sooner than a fixed calendar would allow. A standing series-of-three treats them identically, which is precisely the problem with it.
Why you cannot simply repeat it next week
There are two separate reasons for the waiting interval, and they are often confused.
You need time to know whether it worked
Corticosteroids are not painkillers and not anesthetics. The local anesthetic mixed into the injection can give you relief within hours, but that wears off and tells you very little. The steroid itself acts over roughly two to seven days. Judging the result before about two weeks have passed means judging it before the drug has finished doing whatever it is going to do.
Your endocrine system is still dealing with the last dose
Injected corticosteroid suppresses the hypothalamic–pituitary–adrenal (HPA) axis — the feedback loop governing your body’s own cortisol production. The most noticeable effects, including raised blood sugar and raised blood pressure, typically settle within 48 to 72 hours, but measurable HPA suppression can persist for three weeks or longer. Stacking injections inside that window means stacking systemic steroid effect, which matters a great deal if you are diabetic and matters to everyone eventually.
This is why the interval exists. It is not arbitrary caution, and it is also not a reason to wait six months when four weeks would serve you better.
The type of steroid changes the answer
Not all corticosteroid preparations behave the same way. Longer-acting formulations stay active at the site for longer and can buy more time between injections — but they also produce more prolonged systemic exposure and a higher chance of side effects. Shorter-acting preparations clear faster and have a better safety profile, at the cost of a shorter window of relief.
Which one is appropriate depends on the target, your medical history, and how many injections you have already had. Someone who has had several previous injections is a different prescribing problem from someone having their first. A frequency guideline that ignores which drug was used is not really a guideline.
Joint injections and epidural injections are not the same question
Most of the advice you will find about “how often can I have a cortisone shot” is written about joints — a knee, a shoulder, a hip. That is where the concern about cartilage and tendon damage is most concrete, and where the three-per-year ceiling is most firmly established.
An epidural steroid injection places medication into the epidural space around an inflamed spinal nerve root rather than inside a joint. The cartilage concern does not apply in the same way, but the systemic and cumulative-exposure concerns do, and the response-driven logic above applies with even more force, because an epidural that produces no relief is a strong signal that the nerve root is not the source of your pain. For what the steroid is actually doing once it reaches the root, see why an inflamed nerve root burns, and what an epidural steroid injection actually does.
Injections into the sacroiliac joint, the hip joint, or trigger points each carry their own considerations. Asking “how many can I have” without saying where is asking an unanswerable question.
What should happen instead of a standing schedule
A defensible approach looks like this:
- One injection, then a real assessment. You record what changed, how much, and for how long — not a general impression weeks later.
- The result decides the next step. Good response, consider repeating. Partial response, consider repeating and reconsider the target. No response, stop and re-examine the diagnosis.
- Cumulative exposure is tracked, not guessed. Every previous injection counts, including ones given elsewhere and ones given in a different joint.
- The plan is stated before you start, including what would make us stop.
If you are being told the number before anyone has seen how you respond, you are being scheduled rather than treated.
When “how many more injections” is the wrong question
Sometimes the honest answer is that injections have done what they can, and continuing to ask how many more you can have is asking the wrong question. Depending on the problem, the more useful conversation may be about genicular nerve blocks or genicular radiofrequency ablation for knee pain, viscosupplementation, other nerve blocks, shockwave therapy, spinal cord stimulation, or the metabolic and lifestyle factors that keep inflammatory pain going in the first place.
An injection that keeps being repeated because nothing else was offered is not a treatment plan.
Frequently asked questions
How soon can you have a second cortisone injection?
Rarely sooner than about two weeks, because that is roughly how long it takes to judge whether the first one worked, and because HPA suppression from the first dose can persist for three weeks or longer. Many practitioners prefer six to twelve weeks between injections at the same site. The right interval for you depends on your response, the steroid used, and how many you have already had.
How long do cortisone injections last?
It varies widely — from no useful relief at all to several months. A common range is around two to three months, but duration is not fixed and tends to be shorter with repeated injections at the same site. Duration is one of the things you should be recording, because it is a key input into whether repeating makes sense.
Why do people say cortisone shots are bad for you?
Because repeated steroid exposure at the same site is associated over time with weakening of local tendon, ligament and cartilage tissue, and because systemic absorption raises blood sugar and blood pressure. None of that makes a well-indicated injection a bad idea. It makes an unlimited series of them a bad idea, which is exactly why the decision should be response-driven rather than scheduled.
Do you need to rest after a cortisone injection?
Usually briefly. Most people resume normal activity within a day, but treating the injected area gently for the first day or two is sensible, and strenuous loading of a freshly injected joint or tendon is not. Follow the specific instructions you are given, since they differ by site.
Can you drive after a cortisone injection?
Usually, yes. We do not use sedation, so most patients drive themselves home. The exception is an injection that can leave a limb temporarily weak or numb — an epidural can — in which case wait until that has fully resolved before you drive. Ask before your appointment so you know which applies to you.
How often can you have cortisone injections in your knee?
The usual ceiling quoted for a single joint is three to four a year with at least six weeks between them, and many clinicians prefer longer. If a knee is needing injections at that rate, that is itself information — it is usually the point at which alternatives deserve a serious look rather than another shot.
The bottom line
Three a year is a reasonable ceiling and a poor plan. What should decide how often you are injected is your clinical need and your measured response to the last injection — not a number decided before anyone knew how you would do.
If you have been offered a standing series, or you have had several injections without anyone reassessing whether they are still helping, that is worth a second opinion. Contact the Padda Institute to have the question looked at properly.
This article is general information, not medical advice. Decisions about injection frequency should be made with the physician who has examined you and knows your history.
Dr. Gurpreet Singh Padda, MD, MBA, MHP
