Should I Ice After a Cortisone Shot?

Icing after a cortisone shot is one of the most common instructions patients receive, yet the direct evidence supporting it is surprisingly weak. The only randomized controlled trial that specifically tested ice after a corticosteroid injection, conducted on hand and wrist injections, found no significant difference in pain scores between patients who iced and those who did not. That does not mean ice is harmful or that you should ignore your doctor’s advice, but it does mean the recommendation rests more on general pain-management principles than on cortisone-specific research. Whether icing helps you will depend on why you are reaching for it, how your body reacts to the injection, and where the shot was given.

What Actually Happens After a Cortisone Injection

Cortisone shots deliver a synthetic corticosteroid directly into or near a joint, tendon, or bursa to reduce inflammation. The medication itself is anti-inflammatory, so the idea that you would need ice to control swelling right afterward can seem contradictory. But the injection process involves a needle penetrating tissue, and the steroid solution itself can irritate the surrounding area before it takes full effect. Many patients experience some soreness at the injection site within the first few hours, and a smaller but meaningful fraction experience what clinicians call a “post-injection flare,” a temporary spike in pain that can feel worse than the condition being treated.

In a prospective study of 140 patients receiving musculoskeletal corticosteroid injections, about one in five reported a pain flare after the procedure.¹ A separate study focused on shoulder injections found the rate was even higher, with roughly a third of patients experiencing delayed post-injection pain lasting an average of about four days.² The flare is thought to result from a crystal-induced reaction as the steroid particles settle in the tissue. The good news from that shoulder study is that the flare did not affect the overall outcome of the injection. It is a temporary nuisance, not a sign that something went wrong.

What the Trial Evidence Says About Icing

Given that post-injection pain is common, you would expect a solid body of research on whether ice helps. In reality, the evidence is thin. The most directly relevant study is a randomized controlled trial that assigned patients receiving corticosteroid injections in the hand or wrist to either apply ice alongside over-the-counter painkillers or take painkillers alone. Pain scores were tracked before the injection and for five days afterward. The result was clear: there were no significant differences in pain at any time point between the two groups. In the statistical modeling, ice did not predict pain reduction at all.³

If anything, the numbers trended in a slightly unexpected direction. More patients in the ice group (17 out of 36) experienced a pain increase of at least two points on a ten-point scale compared with the control group (10 out of 32). The study authors did not claim ice made things worse, since the difference was not statistically significant, but it certainly was not evidence that ice helped.

This is a single trial, and it was limited to one anatomical site. No comparable randomized trial has been published for knee, hip, or shoulder injections. So the honest summary is that the only direct test of ice after a cortisone shot found no benefit, but we cannot be certain that finding applies everywhere in the body. Joints with more surrounding soft tissue and different blood supply patterns could respond differently. The absence of evidence is not the same as evidence of absence, but it is worth knowing that the recommendation to ice is largely extrapolated from general injury-management research rather than from cortisone-specific data.

Why Ice Is Still Recommended Anyway

The biological rationale for icing comes from decades of research on soft-tissue injury, not from cortisone studies specifically. In animal models, applying cold after a muscle contusion reduced swelling by decreasing the permeability of tiny blood vessels, essentially making the walls of those vessels less leaky.⁴ Prolonged cooling in similar experiments also reduced inflammation and tissue damage in the area.⁵

These findings are well established, but they describe what happens after a contusion or crush injury, not after a needle puncture delivering anti-inflammatory medication. A cortisone injection causes far less tissue trauma than the injuries used in those experiments. The corticosteroid itself is already doing the anti-inflammatory work that ice would theoretically contribute to. So the mechanism that makes ice helpful after a sprained ankle does not map neatly onto the post-cortisone situation, where the main source of discomfort is often the crystal reaction from the steroid suspension rather than tissue-level swelling.

That said, ice does numb the skin. If you are sore at the injection site, cold application can reduce how much pain you feel even if it is not meaningfully changing the underlying inflammation. This analgesic effect is real and well documented, and it may be the main reason icing after a cortisone shot “works” for the people who report it helps. Recent research on healthy adults found that cold compression produced strong pain relief, with the analgesic effect plateauing at around ten minutes of application.⁶ Beyond 15 minutes, the body’s rebound warming response started to kick in, increasing blood flow back to the area. So if you do decide to ice, shorter bouts of about ten minutes seem to give you the most pain relief without triggering a counterproductive rush of blood back to the site.

How Long to Ice and What to Use

If your doctor tells you to ice and you want to follow that advice, or if you simply find that it makes the site feel better, the practical question becomes how long and with what. Based on the cold compression research mentioned above, ten minutes is the sweet spot for pain relief. You can extend to 15 minutes if you feel you need a bit more, but going longer than that starts to provoke a reflexive increase in blood flow that can undo the vasoconstriction you were trying to achieve. Icing for 20 to 30 minutes straight, as some older guidelines suggest, may actually be too much.

An ice pack wrapped in a thin towel is the standard approach. Do not apply ice directly to skin, especially over a fresh injection site where the skin may already be irritated. A comparison of crushed ice versus a commercial cryotherapy device on knees found that plain ice lowered tissue temperature significantly more than the device, but it was also rated as more painful by patients during the first hour of use.⁷ For post-injection comfort where you are icing for minutes rather than hours, a bag of crushed ice or frozen peas in a cloth works fine. Expensive cryotherapy wraps are not necessary for this purpose.

Rest Matters More Than Ice

If there is one post-injection instruction with solid evidence behind it, it is rest. A randomized trial of patients with inflammatory knee arthritis found that 24 hours of strict bed rest after an intra-articular steroid injection produced greater clinical improvement lasting up to six months compared with patients who were injected and sent home as usual.⁸ The likely explanation is that immobilizing the joint slows the absorption of the steroid out of the tissue, keeping a higher concentration at the target site for longer. Research on cartilage and bone markers supports this idea, showing that post-injection rest affects how quickly the steroid is resorbed.⁹

For athletes and active people, a clinical review recommended one to two days of rest for the affected joint with a gradual increase in activity afterward, balancing the need to keep the steroid concentrated locally against the practical reality that few people can spend a full day in bed.¹⁰ The concern is not just about comfort. Loading a joint immediately after a cortisone injection could accelerate cartilage damage in a tissue that is temporarily more vulnerable, and systemic absorption increases when you are active, meaning more of the steroid enters your bloodstream rather than staying where you want it.

This is where the evidence makes a practical point that goes underappreciated. People worry about whether to ice or not, but the research suggesting rest improves outcomes by months is much stronger than anything we have for icing. If you are going to prioritize one post-injection behavior, resting the joint is the better bet.

Does Lidocaine in the Shot Change Anything?

Many cortisone injections include a local anesthetic like lidocaine mixed with the steroid. When lidocaine is included, you may feel near-immediate numbness at the site that lasts an hour or two, followed by a window of discomfort before the steroid begins working. A double-blind trial testing whether lidocaine reduced pain during and after corticosteroid injection found that the anesthetic did lower pain intensity, suggesting that its numbing effect is real and not just a volume-related phenomenon.¹¹

When lidocaine is part of the injection, the window where icing might help shifts. You probably will not feel much for the first hour or two anyway, and by the time the local anesthetic wears off, the steroid is beginning to take effect. The uncomfortable gap is real but usually short. If your injection did not include a local anesthetic, the first several hours tend to be more uncomfortable, and that is when people most want to reach for ice. Ask your provider whether lidocaine was included so you know what to expect.

When You Should Not Ice

There are situations where icing after a cortisone shot is genuinely a bad idea. If you have Raynaud’s phenomenon or another condition that causes exaggerated blood vessel constriction in response to cold, applying ice could trigger a painful vasospastic episode, especially in the hands or feet. People with peripheral neuropathy who cannot fully feel temperature changes risk frostbite without realizing the ice is too cold. If the injection was deep, such as a hip joint injection performed under imaging guidance, surface ice will not reach the injection site anyway and serves no purpose.

You should also be alert to signs that something beyond a normal flare is happening. Redness and warmth that get worse rather than better over two to three days, fever, or a joint that becomes dramatically more swollen could signal an infection. These are uncommon but serious, and icing them would mask symptoms while the problem worsens. One review of 278 complication cases following corticosteroid injections found treatment errors in about 40% of them, including failures to maintain sterile technique.¹² Infection after a cortisone shot is rare, but when it happens, the symptoms can mimic a flare. If you are unsure, call your provider rather than icing and waiting it out.

Heat as an Alternative

Some patients instinctively reach for a heating pad instead of ice, especially if the injected area feels stiff rather than acutely sore. Heat increases blood flow, which in theory could speed the absorption of the steroid away from the site, the opposite of what you want in the first day or two. Most clinicians advise against heat for the first 24 to 48 hours for that reason. After the initial period, gentle warmth can help with residual stiffness, particularly in larger joints like the knee or shoulder.

The hand-and-wrist trial discussed earlier compared ice to no ice but did not include a heat arm, so there is no direct randomized evidence pitting heat against ice or heat against nothing in the cortisone-injection context. The advice to avoid early heat is based on pharmacokinetic reasoning rather than trial data.

Repeated Injections and the Longer View

A question that often comes up alongside post-injection care is whether repeated cortisone shots cause joint damage. A long-term study tracking patients who received repeated injections into knee joints over periods ranging from four to 15 years found that only two out of 65 cases showed gross deterioration on imaging, and most showed minimal or moderate changes.¹³ That does not mean repeated injections are risk-free, but it pushes back on the common fear that cortisone inevitably destroys cartilage. The relationship between injections and joint health is dose- and frequency-dependent, and your provider will typically space injections months apart to minimize cumulative exposure.

This matters for the icing question because patients who have had multiple injections often develop their own post-injection routine. Some swear by ice, others find it does nothing. Given that the only randomized trial found no measurable benefit, personal comfort is a reasonable guide. If you ice and it makes you feel better, the placebo effect is real pain relief; if you skip it and feel fine, you are not missing a critical step. What the evidence does support is resting the joint, taking over-the-counter pain medication if needed, watching for signs of infection, and following your doctor’s guidance on when to resume activity.

Where the Shot Goes Changes the Equation

Not all cortisone injections are created equal. A shot into a superficial tendon sheath around the wrist involves a small volume of medication close to the skin surface, where ice can actually reach the tissue. An injection deep into the hip joint sits under layers of muscle and fat, and surface ice has no meaningful effect on tissue temperature at that depth. Knee injections fall somewhere in between. The trial showing no benefit from ice was limited to hand and wrist injections, which are among the most superficial sites, making the negative finding even more striking. If ice could not help in a location where cold easily penetrates to the injection site, there is little reason to expect it would work better in deeper joints.

Shoulder injections deserve a special mention because the flare rate appears to be higher there than at other sites, with about a third of patients experiencing post-injection pain lasting several days.² Despite the higher flare rate, no randomized trial has tested ice specifically after shoulder cortisone injections. Patients who experience shoulder flares often find that a sling or simply avoiding overhead movement for a day provides more relief than ice does, but this is clinical observation rather than trial-tested guidance.

Epidural steroid injections for back pain are in a category of their own. These are given under fluoroscopic guidance deep in the spine, and surface icing has no physiological pathway to reach the injection site. Some patients ice the skin entry point for needle-stick soreness, which is fine but is not really about the cortisone at all.