Carpal Tunnel Treatment: Splints, Exercises and Surgery

Most people with mild or moderate carpal tunnel syndrome start with three steps: a neutral wrist splint worn at night, fewer of the tasks that strain the wrist, and a few weeks to see whether symptoms settle. A steroid injection gives faster relief, but the benefit may not last. Surgery to release the transverse carpal ligament, the band that forms the roof of the carpal tunnel, gives the most durable results in the trials we reviewed. Whether mild symptoms clear up unaided is a separate question, covered in whether carpal tunnel goes away on its own.

Treatment follows severity: mild cases start with activity changes, a splint and time

The NHS says surgery may be discussed if symptoms are getting worse and other treatments have not worked, and that surgery usually cures the condition ¹. The order follows from that: non-surgical care first, and surgery when it fails.

This guide follows that order for mild and moderate symptoms. Constant numbness, or weakness in the thumb, is the signal to see a clinician sooner, because in severe cases compression can cause lasting damage ². A clinician usually confirms the diagnosis before choosing treatment, and our guide to how carpal tunnel is diagnosed explains what that involves.

The table is a summary, and each row is explained in its section below. Carpal tunnel syndrome is excessive pressure on the median nerve at the wrist ⁴. Every option aims to reduce that pressure or the symptoms it causes. Typing is not a proven cause: studies have not shown that typing, mouse use or repeated movements cause the condition ⁵. If the pain is in the wrist but the numbness is not, our guide to wrist pain from typing explains other causes.

The flowchart below summarizes the order these sources describe. It is not a protocol for any one person.

Our guide to the symptoms and causes of carpal tunnel syndrome covers how the condition is recognized, and the choice between options depends on how long symptoms have lasted and whether they are getting worse. Those two questions matter more than any single remedy.

A neutral night splint is the usual first step, though the evidence for it is modest

The NHS advises a night splint ⁶, but the trial evidence for splints is weaker than that advice suggests. The 2023 Cochrane review pooled randomized trials of adults with carpal tunnel syndrome. Its finding on symptoms is that splinting may give little or no symptom benefit in the first three months ³.

Some people do report feeling better. The review found that people who used a night-time splint tended to say their overall condition had improved, compared with people who did not use one ⁷. The review treats that finding as uncertain, so it does not establish a measurable gain.

The review also compared how long and how often to wear a splint. One study found that six months of splinting may improve symptoms and function compared with six weeks ⁸. Another found that full-time wear may not improve outcomes compared with night-time wear ⁹. The review does not settle which schedule is best overall, so the schedule is a question to raise with your clinician.

Splints can cause side effects, though they are usually temporary. The review says splinting may cause difficulty falling asleep or transient tingling after removal, and that none of the trials reported serious side effects ¹⁰. If a splint disturbs your sleep, that is worth raising with a clinician rather than simply stopping.

Wear the splint as the NHS describes, at night, and allow time. The NHS says it may take up to six weeks before symptoms start to improve ⁶. A simple written record helps during that period: note each night whether you woke with numbness, how long it lasted, and whether gripping a mug or a bag has become harder. If the splint has not helped by six weeks, that record gives your clinician something concrete to act on, and a clinician can review the options. Which type of brace to buy is a separate question, covered in the question on the best brace for carpal tunnel.

Exercises and hand mobilization have only very weak evidence behind them

A 2012 Cochrane review identified 16 studies of exercise and mobilization, and its conclusion was that the evidence of benefit is limited and very low quality ¹¹. Its search stopped in January 2012 ¹¹, so trials published since then are not included.

An older Cochrane review that dates from 2003 listed yoga among treatments with significant short-term benefit ¹². That is a reason to treat yoga as a possible short-term aid, not as a proven treatment.

People often choose exercise because it is inexpensive and can be done at home. That is a reasonable reason to try it alongside a splint. Judge it by how your symptoms change over several weeks, not by the belief that it repairs the nerve, because the trial evidence does not show that. Keep a note of the exercises you do and what changes, so the question of whether they help is answered from your own record rather than from memory.

We did not verify a source for massage or kinesio taping, so this guide does not rank them. For the evidence on exercise alone, see whether carpal tunnel exercises work.

A steroid injection gives short-term relief, but the benefit does not last

A steroid injection gives short-term relief that the evidence supports. The 2023 Cochrane review pooled nine studies with 639 people ¹³. It found that the injection probably improves symptoms at up to three months, with a standardized mean difference of -0.77 favoring the injection ¹⁴.

Some people also improve without any treatment, so every treatment has to be judged against that background. Spontaneous improvement can occur in up to a third of people ¹⁵. In one trial of 111 people, the risk ratio for needing surgery at one year was 0.84 ¹⁶. That is a relative reduction from a single trial, so it is a weak basis for planning.

The benefit does not last for everyone. In a cohort from a Scottish hand unit, the estimated rate of later carpal tunnel decompression after an injection was 15% at one year and 33% at five years ¹⁷. The figures describe how often injections fail, not how they compare with surgery, because that cohort had no comparison group.

The guidance on the long term is stricter than the trials on the short term. The 2024 AAOS guideline says that strong evidence suggests corticosteroid injection does not provide long-term improvement ¹⁸. The two positions fit together: relief in the first months, but not a lasting cure. Against surgery directly, the 2024 Cochrane review found very uncertain evidence and could not draw conclusions ¹⁹.

Before an injection, ask your clinician how much relief to expect, how long it is likely to last, and what the next step is if it does not work. The 2023 review measured benefit up to three months ¹⁴, so a planned follow-up is the way to find out how it went for you. For more on this question, see whether steroid injections help carpal tunnel.

Oral steroids gave short-term benefit in an old review, and one small trial found no benefit from vitamin B6

Oral steroids were covered in a 2003 Cochrane review, which concluded that they gave significant short-term benefit ¹². The injection evidence above should not be assumed to apply to tablets. Its date matters too: it predates the 2023 splinting review ¹², so treat its conclusions as a starting point rather than current consensus.

We found one randomized, placebo-controlled trial of vitamin B6 that we could verify. It randomized 32 people and found no difference in nerve tests, clinical signs or significant symptoms between pyridoxine and placebo ²⁰. A trial of 32 people cannot rule out a small effect, but it gives no reason to expect a large one.

We did not find a controlled trial of painkillers or diuretics for carpal tunnel syndrome that we could verify, so this guide does not rank them. If pain is the main problem, a clinician can advise on what is suitable for you.

Surgery to release the carpal ligament has the best long-term results

In the Dutch randomized trial, surgery had higher success rates than a night splint at 3 months and at 18 months, though the gap narrowed over time ²¹ and ²². The trial allocated 176 patients to splinting or open surgery ²³. At three months, 79% of surgery patients and 53% of splint patients reported success ²¹. At 18 months the figures were 90% and 75% ²². By then, 41% of the splint group had also had surgery ²⁴, which makes the 18-month gap harder to read.

The 2024 Cochrane review of surgery against non-surgical care found that surgery probably gives a higher rate of clinical improvement than splinting after 6 to 12 months ²⁵. It also found no trial comparing surgery with placebo surgery or no treatment ²⁶. The size of surgery's own effect is therefore not firmly established, even though it outperformed splinting in these trials.

Timing matters because lasting nerve damage is possible. In severe cases, skin sensation can be permanently reduced, and compression can cause muscle wasting at the base of the thumb ². That is the main reason to act on persistent symptoms rather than waiting indefinitely.

Recurrence after a first operation appears uncommon in the one series we found with long follow-up. Among the patients who had their first operation in that department, none had a recurrence of carpal tunnel syndrome ²⁷. At long-term follow-up, two-thirds had no symptoms, and one-third had mild persisting numbness ²⁸. Those results come from a single unit and from patients' own reports by telephone. For detail, see whether carpal tunnel can come back after surgery.

Recovery is slow. A systematic review of nerve recovery after release found that grip and pinch strength did not show marked recovery until 5 to 7 months after surgery ²⁹. For the timeline, see how long carpal tunnel surgery takes to recover from.

The trials cannot tell you how surgery compares with doing nothing, because no trial tested that. Long-term data come mostly from small single-unit series, such as the one above. We have not verified a current price for surgery in any country, so this guide does not quote one.

Choosing between the options depends on how fast you need relief and how long it must last

The options trade speed against durability. If you need relief quickly, an injection has moderate-certainty evidence for short-term benefit, with a standardized mean difference of -0.77 at up to three months ¹⁴. If you need a result that lasts, surgery looked better at 18 months in the Dutch trial ²², though that comparison is weaker because some splint patients had surgery by then. Splints sit at the least certain end, with benefits that are hard to pin down.

The sequence these sources describe is therefore a stepped one. Most people try a splint and activity changes first. If symptoms persist, or get worse, the choice narrows to an injection for quicker relief or surgery for a lasting result, and the trial evidence on injection against surgery is too uncertain to rank them ¹⁹. A clinician who knows your nerve test results and your work is best placed to weigh these trade-offs with you.

You can keep working during treatment by moving typing to your voice

In a laboratory study of healthy volunteers, typing raised the pressure inside the carpal tunnel compared with holding the same hand posture still ³⁰. So the aim while you recover is fewer keystrokes, not only a better chair.

Speech input can take over drafting, email and messages, so the wrist does less work. It does not remove all typing, because editing, short replies and code still need hands. A practical sequence looks like this:

1. Move first drafts to speech. Start emails, notes and documents by dictating, and leave the fine editing for later.
2. Keep the keyboard for short fixes. Use shortcuts for quick corrections rather than long typed passages.
3. Break up the day. Our guide to how often to take typing breaks explains how to space them.
4. Keep the splint on at night as your clinician has advised, and review the plan if symptoms change.

Our guide to dictation for RSI and carpal tunnel sets out a staged plan in more detail, and typing ergonomics covers desk setup. For the umbrella term, see repetitive strain injury, and for the typing side of it, our guide to RSI from typing.

Yap, which we make, is an on-device dictation app for Mac and Windows. If you want to start for free, the built-in tools may be enough: see voice to text on Mac and dictation on Windows. How to choose a dictation app compares the options, and best dictation software covers the wider field.

If you need a formal adjustment at work, our guide to dictation as a workplace accommodation explains how that works. If numbness is constant or the thumb is weakening, see a clinician rather than waiting, because the long-term effects described above are the reason. For the wider subject, see our overview of speech to text for accessibility.