Cortisone shots fail in the knee more often than most people realize, and the reasons range from the needle landing in the wrong spot to the pain actually originating from a completely different structure than the joint itself. Roughly half of people who receive an intra-articular corticosteroid injection for knee osteoarthritis meet clinical criteria as “responders” at follow-up, which means the other half get little or no meaningful relief. Understanding why your shot didn’t work is the first step toward figuring out what will.
The Injection May Not Have Reached the Joint
This is the most straightforward and probably the most underappreciated reason a cortisone shot can fail: the medication never actually made it into the joint space. When a doctor injects your knee using anatomical landmarks alone, they’re essentially making an educated guess about where the needle tip ends up. A systematic review of randomized controlled trials found that ultrasound-guided injections were more accurate than blind injections in every study examined, while blind injections had widely inconsistent accuracy rates depending on the approach used.1PubMed Central. Ultrasound-Guided Knee Injections Are More Accurate Than Blind Injections: A Systematic Review of Randomized Controlled Trials Another review put numbers on the gap: ultrasound-guided knee injections hit the joint space about 96% of the time compared with about 78% for anatomically guided injections.2PubMed Central. Clinical utility of ultrasound guidance for intra-articular knee injections: a review
That accuracy gap translates directly into clinical results. A systematic review of predictors of response found that ultrasound-guided injection was associated with a 42% greater reduction in pain scores at two weeks, more than double the responder rate, and over a third longer duration of symptom relief compared with blind injection.3PubMed Central. Predictors of response to intra-articular steroid injections in knee osteoarthritis—a systematic review A meta-analysis of randomized trials in knee osteoarthritis confirmed that patients receiving ultrasound-guided injections reported better pain relief and function at follow-up.4PubMed Central. Comparison of ultrasound guidance with landmark guidance for symptomatic benefits in knee, hip and hand osteoarthritis: Systematic review and meta-analysis of randomised controlled trials
If your injection was done without imaging guidance, a missed joint space is a real possibility. This is especially true if you have a large body habitus, unusual anatomy, or significant swelling that distorts the landmarks your doctor was relying on. Ask whether your next injection can be done under ultrasound. It adds a few minutes to the procedure but substantially changes the odds.
The Pain May Not Be Coming From Inside the Joint
Cortisone injected into the knee joint treats inflammation inside that joint. If the actual source of your pain is a structure outside the joint, the medication is in the wrong neighborhood entirely. One of the most common culprits is pes anserine bursitis, an inflammation of a bursa on the inner side of the knee just below the joint line. A study of 245 patients diagnosed with knee osteoarthritis found that 71% of them also had pes anserine bursitis.5International Journal of Research in Medical Sciences. Frequency of pes anserine bursitis in patients with knee osteoarthritis That’s a striking number. It means a large portion of people with knee osteoarthritis have a second, separate source of pain that an intra-articular cortisone shot wouldn’t touch.
Other structures that can mimic or contribute to knee-joint pain include the iliotibial band, Baker’s cysts, meniscal tears, and ligament injuries. If your pain is predominantly on the inner side of the knee below the joint line, or if it feels more like a burning ache along the tendons rather than a deep joint ache, the cortisone may have gone to the right place but your pain wasn’t coming from there. A targeted physical exam or an ultrasound can help sort this out.
There’s also a related misconception worth addressing. Many people assume that if you can see swelling in the knee, cortisone should work especially well because there’s clearly inflammation to suppress. A pilot study tested exactly this, comparing cortisone injection outcomes in osteoarthritis patients with and without visible effusion on ultrasound. The researchers expected effusion to predict a bigger drop in pain. It didn’t: the difference in pain reduction between those with and without effusion was neither clinically meaningful nor statistically significant.6PubMed Central. Response to intra-articular cortisone injections in knee osteoarthritis patients with and without effusion on ultrasound: A pilot study So a visibly swollen knee doesn’t guarantee you’ll respond, and a knee without obvious swelling isn’t necessarily a lost cause.
When Pain Has Moved Into the Nervous System
In some people with longstanding knee osteoarthritis, the pain processing itself has changed. The nervous system has become sensitized, amplifying pain signals so that even minor stimulation of the joint produces outsized discomfort. This phenomenon, called central sensitization, means the problem isn’t just in the knee anymore; the brain and spinal cord are independently driving the experience of pain. A cortisone shot that reduces inflammation at the joint level can’t fix a nervous system that’s turned up the volume.
A study of older knee osteoarthritis patients found that the presence of central sensitization significantly limited the clinical response to both intra-articular injections and radiofrequency nerve ablation.7PubMed. Does central sensitization affect responses to genicular radiofrequency and intra-articular injection in elderly knee osteoarthritis patients? The researchers recommended routine screening for central sensitization before deciding on a treatment approach. Clues that central sensitization may be involved in your case include widespread pain sensitivity beyond the knee, pain that seems disproportionate to what imaging shows, heightened sensitivity to touch or pressure, and difficulty sleeping due to pain. If these ring true, you may benefit more from treatments that target pain processing, such as certain medications, cognitive behavioral therapy, or graded exercise programs, rather than another injection.
Some People’s Cells Are Resistant to Steroids
This is less well known but supported by a growing body of research: not everyone’s cells respond to corticosteroids the same way. The glucocorticoid receptor comes in multiple forms, and one variant in particular can block the normal anti-inflammatory effect. A systematic review examining steroid resistance across multiple inflammatory conditions found that elevated levels of this receptor variant were consistently associated with poor response to steroid treatment. This held across asthma, nasal polyps, ulcerative colitis, Crohn’s disease, and rheumatoid arthritis.8PubMed Central. Association between glucocorticoid receptor beta and steroid resistance: A systematic review
The glucocorticoid receptor is also subject to various chemical modifications inside cells that can alter how well it functions, and the mix of receptor forms expressed varies from person to person and even from tissue to tissue. This means there’s genuine biological variability in how strongly your joint tissue responds to cortisone. If you’ve had multiple cortisone injections in different joints and never experienced meaningful relief from any of them, cellular steroid resistance is worth considering as a factor. Unfortunately, there’s no routine clinical test for this yet, but it does explain why some people simply don’t get the response that others do from the same medication.
The Placebo Effect of Knee Injections Is Surprisingly Large
Here’s something that may reframe how you think about your shot. When researchers run clinical trials on knee injections, the placebo group receives a saline injection rather than the active drug. And those saline injections produce remarkably large improvements. A meta-analysis found that placebo knee injections resulted in clinically meaningful improvements in pain, stiffness, and function that persisted up to six months, with a responder rate of about 52%.9PubMed Central. The Long-Lasting Effects of “Placebo Injections” in Knee Osteoarthritis: A Meta-Analysis
A separate analysis focused on placebo responses in biologic injection trials found that placebo injections produced the greatest pain reduction at one month and were significantly more effective at reducing pain than oral placebos.10PubMed Central. Placebo Effect Sizes in Clinical Trials of Knee Osteoarthritis Using Intra-Articular Injections of Biologic Agents In other words, the ritual of the injection itself, the expectation of relief, the attention from a medical provider, and the simple act of having a needle enter the joint all produce real, measurable changes in pain and function.
A large systematic review and meta-analysis of placebo responses in knee injection trials confirmed that about half of placebo recipients qualified as responders at the one-, three-, and six-month marks, with pain, stiffness, and function all improving by clinically meaningful amounts.11PubMed Central. Placebo response to intra-articular injections in knee osteoarthritis: magnitude, evolution over time, and influencing factors. A systematic review and meta-analysis with meta-regression The improvement from an actual cortisone shot, then, has to be measured against this already-substantial placebo baseline. In some patients the cortisone provides a meaningful benefit on top of the placebo effect, but in others the drug itself may add little. If a previous injection seemed to help for a week or two and then faded, part of what you experienced may have been the placebo component wearing off rather than the cortisone failing.
The Relief Was Real but It Wore Off
Cortisone is not a cure. It’s a temporary anti-inflammatory, and its pain-relieving effect decreases over time. A study comparing two common steroid formulations, methylprednisolone and triamcinolone, in knee osteoarthritis found that both produced a significant drop in pain scores shortly after injection, but the effect steadily declined over the 24-week follow-up period.12PubMed Central. Compared Efficacy of Intra-articular Injection of Methylprednisolone and Triamcinolone Both formulations performed similarly, so your doctor’s choice of one steroid over another probably didn’t make a difference. What matters more is understanding that the effect is inherently time-limited.
For some people, the window of relief lasts a few months. For others, it’s measured in weeks. If your shot “worked” for three weeks and then your pain returned to baseline, that’s actually a common trajectory, not a failed injection. It means the cortisone did suppress inflammation temporarily, but the underlying disease reasserted itself. The question then becomes whether repeated injections are the right strategy, whether physical therapy during the pain-free window can build strength that outlasts the drug, or whether it’s time to consider other options.
What you do immediately after the injection also plays a role. Expert recommendations suggest one to two days of rest for the affected joint following a corticosteroid injection, with a gradual return to activity.13PubMed Central. One to Two Days of Rest Is Recommended Before Returning to Sport After Intra-Articular Corticosteroid Injection in the High-Level Athlete The rationale is both practical and biological: rest may help maximize the beneficial effects and reduce how quickly the steroid is absorbed out of the joint. If you walked out of the office and immediately resumed heavy activity, the cortisone may have been cleared from the joint before it had a full chance to work.
What Repeated Injections Do to Cartilage
If you’re thinking about getting another shot because the first one didn’t work, there’s a conversation worth having with your doctor about cartilage health. A systematic review of lab and animal studies found that corticosteroids displayed dose-dependent harmful effects on cartilage. At lower doses the damage was minimal, but at higher doses or higher cumulative totals, corticosteroids were associated with significant cartilage damage and death of the cells that maintain cartilage.14PubMed Central. The Effect of Intra-articular Corticosteroids on Articular Cartilage: A Systematic Review
Human imaging data points in the same direction. A meta-analysis of studies using MRI and X-ray to track cartilage changes over time found that corticosteroid-injected knees had roughly double the odds of cartilage worsening compared with controls.15PubMed Central. Longitudinal MRI-defined Cartilage Loss and Radiographic Joint Space Narrowing Following Intra-Articular Corticosteroid Injection for Knee Osteoarthritis: A Systematic Review and Meta-analysis This doesn’t mean a single injection will destroy your knee. But it does mean that chasing relief with repeated injections in a knee that isn’t responding well may come at a structural cost. If the first shot didn’t help, a second one of the same thing is unlikely to produce a different result, and it may accelerate the cartilage loss that’s driving your osteoarthritis in the first place.
Combining Cortisone With Other Treatments
One approach that has shown some promise for patients who get incomplete relief from cortisone alone is combining the steroid injection with platelet-rich plasma, or PRP. A clinical trial comparing cortisone alone with cortisone plus PRP found that the combination group had significantly greater improvement in pain and stiffness scores.16Journal of University College of Medicine and Dentistry. Efficacy of Intra-Articular Platelet-Rich Plasma plus Corticosteroid Versus Corticosteroid Alone in Knee Osteoarthritis: A Clinical Interventional Study The improvement in functional limitation wasn’t significantly different between groups, and the overall score difference narrowly missed statistical significance, so this isn’t a slam-dunk. But it suggests that adding a biologic component may enhance the anti-inflammatory window of cortisone in some patients.
PRP is far from the only alternative or complement. For people whose cortisone shot failed because of central sensitization, a more productive route might be a structured exercise program combined with pain neuroscience education. For those whose pain turned out to be from pes anserine bursitis, a targeted injection to that specific bursa, different physical therapy exercises, or a simple change in gait mechanics may do what the intra-articular shot couldn’t. The answer to “what’s next” depends entirely on why the first shot didn’t work.
Body Weight and Structural Severity
Two factors that reliably influence outcomes across nearly every knee osteoarthritis treatment are body weight and how far the disease has progressed. Higher body weight means more mechanical load on the joint, a larger volume of tissue for the drug to diffuse through, and often more pervasive low-grade inflammation driven by fat tissue itself. Patients with advanced structural disease, where the cartilage is extensively worn and bone-on-bone contact has begun, tend to get less out of conservative treatments in general, because there’s less functioning joint left to calm down.
If your knee X-ray shows severe narrowing or bone spurs and you’re carrying significant extra weight, these factors may have stacked the odds against a cortisone injection from the start. That doesn’t mean the injection was the wrong call, since even partial or temporary relief has value, especially if it enables you to start exercising. But it does mean your expectations and next steps should be calibrated to the reality of how advanced the disease is. For people with late-stage osteoarthritis who don’t respond to injections, the conversation often shifts toward surgical options like partial or total knee replacement.
When to Suspect the Diagnosis Itself Is Wrong
If your cortisone shot produced zero relief, not even a few days of reduced inflammation, it’s worth questioning whether osteoarthritis is the right diagnosis. Several other conditions can produce knee pain and even look like arthritis on imaging. Referred pain from the hip is a classic: hip osteoarthritis frequently presents as knee pain, and no amount of cortisone in the knee will fix a hip problem. Inflammatory arthritis conditions like gout or pseudogout can flare and subside in ways that mimic osteoarthritis, and while cortisone can help during a flare, it won’t address the underlying crystal deposition. Nerve-related pain from lumbar spine issues can radiate to the knee and mimic a joint problem entirely.
Imaging can be misleading here too. Many people over 50 have some degree of osteoarthritis visible on X-ray or MRI, but that doesn’t mean the arthritis is causing their current pain. Structural findings on imaging don’t always correlate with symptoms. A person can have a knee that looks terrible on X-ray but functions reasonably well, while someone with mild imaging findings can be in significant pain from a soft-tissue problem that the X-ray wasn’t designed to detect. If your cortisone shot produced no response at all and you haven’t had a thorough clinical evaluation beyond imaging, getting a fresh set of eyes on the problem may be more productive than trying another injection.