How Long After a Cortisone Shot Can I Have an MRI?

Most radiologists and pain specialists recommend waiting at least four days after a cortisone shot before having an MRI of the same area, though some clinicians prefer a buffer of one to two weeks depending on the injection site and the type of steroid used. The concern is not safety in the way most patients imagine it: the cortisone will not react with the MRI’s magnetic field or cause tissue damage. The problem is that a recent injection can change what the MRI looks like, potentially leading a radiologist to misread the images. How long those artifacts persist depends on where the shot was given, what formulation was used, and what the MRI is trying to evaluate.

What a Recent Injection Does to MRI Images

When a cortisone shot is delivered into a joint, a bursa, or the epidural space, the needle and the injected fluid temporarily alter the local tissue environment. Air can be introduced along the needle track. The liquid component of the injection, which includes the steroid suspension plus a local anesthetic and sometimes saline, pools in the space and shows up as extra fluid on the scan. Soft tissues disturbed by the needle can appear swollen or inflamed. None of these changes reflect the underlying condition the MRI is meant to evaluate. They are artifacts of the procedure itself, and a radiologist who does not know about the recent injection could mistake them for signs of disease, injury, or structural damage.

A study tracking patients after epidural steroid injections found that over half of those scanned within roughly 24 hours showed abnormal MRI changes, including epidural air, fluid collections, and visible needle tracks. At two days after injection, about a third still showed artifacts. By three days, a quarter did. But by four days, none of the scanned patients showed injection-related abnormalities on their images.¹ Epidural air was the single most common artifact, appearing in roughly four out of ten patients scanned within the first three days.

This four-day threshold is one of the clearest data points available, but it comes from epidural injections in the spine. Joint injections in the shoulder, knee, or hip introduce their own quirks, and some clinicians set a longer wait for those areas.

Shoulder and Joint Injections Can Create a Longer Window of Confusion

Shoulder injections are a particularly well-documented source of MRI misinterpretation. When corticosteroid is injected into the subacromial bursa, the fluid can seep into surrounding structures and create signal changes on MRI that look remarkably like tendon tears. Research on shoulder injections has led to a general recommendation of waiting at least three days before scanning, based on the observation that bursal and extrabursal fluid typically resolves or decreases within that window.²

But three days may not always be enough. In one well-known case report, an MRI obtained seven weeks after a subacromial corticosteroid injection showed what appeared to be a full-thickness tear of the supraspinatus tendon, a finding that would typically call for surgical evaluation. The patient went on to have arthroscopic surgery, and the surgeon found the rotator cuff tendons completely intact and normal in appearance. The injection had apparently penetrated the tendon and caused a lingering signal change that mimicked a tear on imaging.³

Seven weeks is an unusually long artifact. Most clinicians would not expect injection-related changes to persist that long, and this case likely involved the needle directly traumatizing the tendon. Still, it illustrates why the stakes of timing are higher than most patients realize. A misread MRI can lead to unnecessary surgery, additional testing, or treatment of a problem that does not actually exist.

Why the Type of Steroid Matters

Not all cortisone shots use the same drug. The formulation affects how quickly the steroid is absorbed and cleared from the injection site, which in turn affects how long it might interfere with imaging.

Corticosteroids used in injections fall into two broad categories. Particulate steroids, such as methylprednisolone acetate, triamcinolone acetonide, and betamethasone acetate, are suspensions of tiny crystite particles. They are designed to dissolve slowly, forming a depot in the tissue that releases the active drug over days to weeks. Because of their larger particle size and tendency to aggregate, they linger at the injection site longer and produce a longer duration of pain relief. Non-particulate steroids like dexamethasone sodium phosphate dissolve readily in water, are absorbed quickly by cells, and clear the area faster.⁴

From an MRI timing perspective, this distinction is relevant. A particulate steroid depot sitting in a joint or epidural space may be visible on imaging for longer than a quickly-absorbed non-particulate formulation. If you are unsure which steroid was used in your injection, the prescribing physician or the facility where the injection was performed can tell you. In practice, triamcinolone and methylprednisolone are among the most commonly used formulations for musculoskeletal injections, and both are particulate types with longer local residence times.

How Cortisone Can Mask What the MRI Is Looking For

The timing question has a second dimension that goes beyond fluid artifacts and needle tracks. Cortisone is a powerful anti-inflammatory, and inflammation is one of the things MRI is very good at detecting. If the cortisone is doing its job and suppressing inflammation in the area, an MRI taken while the drug is actively working may underrepresent the severity of the underlying condition. The scan could look deceptively calm.

This effect has been directly measured in arthritic knees. When researchers performed MRI before and at intervals after intra-articular methylprednisolone injection, they found that the rate of synovial membrane enhancement on gadolinium-enhanced MRI decreased across all treated knees within the first week after injection and stayed low throughout the period of clinical remission, only gradually increasing as the steroid wore off.⁵ In other words, the MRI faithfully reflected the reduced inflammation, but that reduced inflammation was pharmaceutical, not natural. A clinician evaluating that scan without context might underestimate how actively inflamed the joint is when not being treated.

A similar phenomenon has been observed with high-dose intravenous steroids in neurological conditions. Patients with acute demyelinating brain lesions showed complete or significant suppression of gadolinium enhancement on MRI after a short course of high-dose steroids.⁶ And in spinal cord sarcoidosis, corticosteroid therapy led to dramatic reductions in swelling and enhancement visible on MRI.⁷

These are intravenous and systemic steroid treatments rather than local injections, so the effect is more pronounced. But the principle applies on a smaller scale to local injections too. If your doctor ordered both the cortisone shot and the MRI as part of a diagnostic workup, the timing between them matters not only because of injection artifacts but because the steroid’s therapeutic effect can temporarily mute the very inflammation the MRI is supposed to capture.

When the MRI Is Ordered After the Shot, Not Before

In an ideal diagnostic sequence, the MRI would come first and the cortisone shot afterward. The scan captures the unaltered state of the tissue, the physician uses that information to decide on treatment, and the injection follows. But real-world medicine does not always work in that order. Sometimes a patient gets an injection for acute pain relief and an MRI is scheduled later, or the injection was given at a different facility and the MRI was already on the books. Sometimes the cortisone shot is being used diagnostically: if the injection into a specific joint relieves your pain, that helps confirm the joint as the source of the problem, and the follow-up MRI is intended to look at structural detail.

When the MRI follows the injection rather than preceding it, the waiting period becomes more important and harder to generalize. Four days clears the acute artifacts from epidural injections. Three days clears most of the fluid from shoulder injections. But the anti-inflammatory masking effect can persist for weeks in a joint, and some structural signal changes, as the rotator cuff case demonstrated, can linger even longer. For this reason, many orthopedic surgeons and radiologists prefer a wait of two to four weeks after an intra-articular cortisone injection before performing an MRI that is supposed to guide surgical decision-making. The evidence is not set in stone for every scenario, which is why individual recommendations vary.

Does the Steroid Interfere with Gadolinium Contrast?

Some MRI scans use an intravenous or intra-articular injection of gadolinium-based contrast to improve image quality. If you are having a contrast-enhanced MRI, you might wonder whether the residual cortisone in the tissue interacts with gadolinium in a way that distorts the scan.

Phantom studies, which test signal behavior in controlled lab conditions rather than in living patients, have looked at exactly this question. Researchers measured gadolinium signal intensity with and without the addition of steroids and local anesthetics. At the standard MRI field strength of 1.5 Tesla, the addition of steroids and anesthetics had no impact on the gadolinium signal curves. At 3.0 Tesla, the effect was minimal.⁸ So from a pure chemistry standpoint, the steroid itself does not meaningfully alter how gadolinium behaves in the scan. The concern remains the tissue-level changes already discussed: the extra fluid, the inflammation suppression, the soft tissue disruption. Those are biological effects, not chemical interactions with contrast agents.

Practical Guidelines by Injection Site

Because the evidence comes from studies of different body regions, there is no single universal recommendation. Here is a rough guide based on published findings and common clinical practice:

Epidural (spine): At least four days, based on the finding that injection artifacts disappear by day four after epidural steroid injections.
Shoulder (subacromial): At least three days to clear injection fluid, but two to four weeks if the MRI is being used to assess for rotator cuff tears or guide surgical planning.
Knee or hip (intra-articular): Many clinicians recommend two to four weeks if the goal is to evaluate the degree of synovial inflammation or structural damage without the confounding effect of active steroid suppression.
Soft tissue or tendon sheath: At least a week is a common informal recommendation, though published evidence specifically for these sites is thinner.

These are not rigid rules from a unified guideline. They are ballpark figures drawn from the available research and clinical convention. Your radiologist or ordering physician may set a different interval based on the specific clinical question the MRI is supposed to answer. If the scan is looking for a fracture or a large structural tear, a few days may be enough because those findings are not easily masked by steroid effects. If the scan is trying to quantify inflammation or distinguish a partial tendon tear from post-injection signal changes, a longer wait is warranted.

What You Should Tell the Radiologist

Regardless of how long you wait, always inform the MRI facility that you have had a recent cortisone injection. Ideally, provide the date, the injection site, and the name of the steroid used. This information gets passed to the radiologist who reads the images, and it can prevent a misdiagnosis. A fluid collection that looks like an abscess might be recognized as residual injectate. A signal change in a tendon that suggests a tear might be flagged as potentially injection-related rather than surgical.

The radiologist’s interpretation is only as good as the clinical context they are given. A 2010 case report on shoulder injection artifacts specifically emphasized that MRI findings soon after corticosteroid injection should be interpreted with caution, because signal changes can convincingly mimic pathology.³ A radiologist who knows about the injection will apply that caution. One who does not may read the images at face value.

When Waiting Is Not Realistic

There are situations where clinical urgency makes a long wait impractical. If you have a suspected spinal cord compression, a rapidly worsening neurological deficit, or symptoms suggesting infection at the injection site, the MRI should not be delayed just because you had a cortisone shot recently. In emergencies, getting the scan is more important than getting a perfect scan. The radiologist can account for the known limitations, and any ambiguous findings can be followed up with repeat imaging later.

Similarly, if your MRI appointment was already scheduled and difficult to rebook, and the injection was given into a completely different body region, there is no reason to delay. A cortisone shot in the knee does not affect the quality of a brain MRI or a lumbar spine scan at a different spinal level. The artifacts are local to the injection site. The systemic anti-inflammatory effects of a local injection are generally too mild to meaningfully change imaging of a distant area.

The Cascade Problem with Poorly Timed Imaging

One underappreciated reason to get the timing right is the risk of a diagnostic cascade. When an MRI produces ambiguous or misleading findings, the typical response is more testing: repeat MRIs, CT scans, additional specialist referrals, and sometimes invasive procedures like arthroscopy or biopsy. Research on early or unnecessary MRI for low back pain has shown that premature imaging is associated with a large and sustained increase in downstream medical services and costs, with patients who received early MRI being many times more likely to undergo additional diagnostic procedures and invasive treatments over the following months.⁹

That study was about imaging that was premature for clinical reasons, not about post-injection timing specifically. But the principle transfers. An MRI contaminated by injection artifacts can generate the same kind of ambiguity that triggers more testing. Getting a clean scan the first time, by waiting an appropriate interval, can spare you a frustrating cycle of follow-ups and false alarms. If you have already waited weeks for an MRI slot, waiting a few more days to ensure the images are interpretable is usually worth it.