Cervical radiculopathy: the exercises that work
Neck pain running down the arm, sometimes with pins and needles: cervical radiculopathy usually runs a favourable course, but slowly. The good news is that active treatment beats waiting from the outset. Here are the exercises the research supports (neural mobilisation, deep flexor strengthening), with their dosages, and an honest account of what each technique is worth.
📝 In brief
- Yes, exercise acts first on pain. A meta-analysis of 10 randomised trials (871 patients) shows that exercise, alone or combined with conventional treatment, significantly reduces the pain of cervical radiculopathy (SMD = -0.89; 95 % CI: -1.34 to -0.44). The gain in function measured on the Neck Disability Index (-3.60 points), on the other hand, remains below the threshold of clinical relevance (MCID ≈ 5 to 7.5 points) and the authors judge the level of evidence to be low (« may help ») 2.
- From the early phase, moving beats waiting. In 205 patients with radiculopathy of less than a month, physiotherapy with home exercises (6 weeks) reduced neck and arm pain substantially compared with a wait-and-see approach (around -14 mm on the neck VAS, -17 mm with collar + rest) 4.
- The prognosis is favourable but slow. Under conservative treatment, around 55 % of patients report themselves recovered at 6 and at 12 months, with most of the improvement occurring in the first few months. A longer duration of symptoms increases the risk of poor recovery, whereas the presence of paraesthesia reduces it 1.
- It is the active programme that counts, not any particular manual technique. In a blinded trial (36 patients), manual therapy and exercise improve pain and limitations both statistically and clinically significantly at 4 and 8 weeks; adding techniques supposed to « open » the foramen adds nothing. Note: both arms were active protocols (no control group) and follow-up does not go beyond 8 weeks 5.
- Cervical traction is not essential. In 81 patients with cervical radiculopathy, adding intermittent mechanical traction to a manual therapy + exercise programme brings no additional benefit for pain or disability 7 ; a 2025 network meta-analysis judges its value still uncertain 6.
- A structured physiotherapy programme should precede any decision on ACDF. At 2 years, physiotherapy alone and anterior cervical decompression and fusion (ACDF) followed by physiotherapy do not differ on functional outcomes (range of motion, muscle endurance, hand function; P = 0.17 to 0.91): the comparison being on function, not pain, and on a small sample of operated radiculopathies 3.
🎯 Understanding it in two minutes
Cervical radiculopathy, also called brachial neuralgia, corresponds to irritation or compression of a nerve root as it leaves the cervical spine. The nerve concerned then runs down the arm: that is why pain, born in the neck, travels down into the shoulder, the forearm or the hand, often with pins and needles, a sensation of « electric shock » or a loss of strength. The cause is most often a disc herniation or cervical osteoarthritis reducing the space through which the root passes.
Good news: under conservative treatment (without surgery), the course is broadly favourable, but slow. Around 55 % of people report themselves recovered at 6 months as at 12 months, with most of the improvement happening in the first few months 1. Two useful qualifications: symptoms that have lasted a long time increase the risk of incomplete recovery, whereas the presence of pins and needles is rather a favourable sign.
Who this programme is for
This exercise programme is for people whose cervical radiculopathy has already been diagnosed and for whom active treatment has been judged suitable. It does not replace the initial examination: it is that assessment which rules out the situations calling for prompt medical advice (a loss of strength that is setting in, significant disturbances, red flags).
If you do not yet have a diagnosis, or if you want to understand the causes, the investigations and the warning signs, start with our full article on cervical radiculopathy. This article focuses on one question only : which exercises to do, and how to dose them.
The guiding principle: reloading, gradually
The thread running through the whole programme comes down to one idea: giving the neck and arm a workload again, in stages, without stirring up lasting pain. We are looking neither for strict rest nor for exercise at any price, but for controlled progression where intensity rises as symptoms allow.
This active logic is supported by the data. A meta-analysis of 10 randomised trials (871 patients) shows that exercise, alone or combined with conventional treatment, significantly reduces pain, with a large effect size (SMD = −0.89; 95 % CI: −1.34 to −0.44) 2. We must nonetheless stay honest about the functional side: on the Neck Disability Index, the mean gain (−3.60 points) is statistically significant but remains below the threshold of clinical relevance (generally put at 5–7.5 points), and the authors themselves describe the level of evidence as low (« may help »). In other words: exercise clearly brings pain down, but its effect on perceived disability is more modest than we would like to be able to claim.
The right exercise is not the one that avoids pain at all costs, it is the one that restores confidence in movement.
Another point to bear in mind: exercise does not compete with the other active techniques. It combines readily with manual therapy and neural mobilisation, and several trials show that what counts is a well-conducted active programme rather than any isolated technique. Finally, when surgery comes into question (in radiculopathy being considered for cervical fusion), the data suggest putting a structured rehabilitation programme first: at two years, function was no better after surgery than after a well-conducted physiotherapy programme 3.
Key points
- Cervical radiculopathy most often heals without surgery, but slowly : be patient and consistent.
- Exercise does reduce pain ; its effect on perceived disability is real but more modest (low-level evidence).
- The principle is progressive loading, not rest, and not pain at all costs.
- The winning approach is active and multimodal : exercise combines with other techniques, it does not compete with them.
- No diagnosis yet? Start with the full article.
💪 The exercise programme, step by step
First, some good news: under conservative treatment, cervical radiculopathy most often moves towards improvement. In a large follow-up study, around 55 % of people report themselves recovered at 6 as at 12 months, with most of the progress occurring in the first few months 1. It is slow, but the slope is in the right direction, and an active programme accompanies that natural movement rather than simply waiting for it. To understand the mechanism and the warning signs, see the full article on cervical radiculopathy; here we focus on the doing.
Key points
- Moving reduces pain: exercise, alone or combined with other treatments, significantly reduces arm and neck pain 2.
- Progression happens in phases, respecting an acceptable level of pain, never out of fear of movement.
- The winning approach is active and multimodal: exercise combines with manual therapy and neural mobilisation, without setting them against each other.
- A structured programme deserves to be seen through before considering decompression and fusion surgery (ACDF) for cervical radiculopathy 3.
The acceptable-pain rule
The thread running through the whole programme: an exercise may stir up tolerable discomfort (around 3–4/10), provided it returns to its baseline level within the following hours and no symptom travels further down the arm. Pain that worsens, spreads towards the hand or lasts until the next day is a signal to do less that day. This rule is a marker of clinical common sense, not a value validated by trial: adapt it with the professional following you.
Phase 1: Settle things and get moving again (weeks 0 to 2)
In the early phase, the aim is not to push but to break the immobility and the fear of moving. In patients whose symptoms had lasted less than a month, six weeks of physiotherapy with home exercises reduced neck and arm pain markedly more than a wait-and-see approach 4.
- Gentle chin retractions (chin drawing back, « double chin »): 2 sets of 10, slow tempo, twice a day.
- Shoulder and thoracic spine mobility (shoulder rolls, seated opening) to unload the neck: 2 × 10, 1–2 times a day.
- Neural glides for the nerve concerned, within a comfortable range, without stretching: 1–2 × 10 slow repetitions, as long as they do not increase the pins and needles.
We stay below the acceptable-pain threshold. You move on when the basic movements can be done without a marked spike of pain.
Phase 2: Mobility, control and neurodynamics (weeks 2 to 6)
We widen the range and wake up the deep stabilisers of the neck.
- Active cervical range of motion (flexion/extension, rotations) within the pain-free range: 2 × 8–10, daily.
- Gentle isometric strengthening (hand resisting without movement, in all 4 directions): 3 × 8–10 s holds, submaximal contraction, once a day.
- Scapular control (retractions, light « Y-T-W »): 2–3 × 10, 3–4 times a week.
- Progressive neural glides, with a slightly larger range as tolerated.
This kind of programme reduces pain and limitations at 4 and 8 weeks 5; to be clear, that trial compared two active protocols with each other, without a control group, and measured only the short term: it shows real improvement, not superiority over the natural course.
Phase 3: Strength and endurance (weeks 6 to 12)
We load more to build robustness: neck, shoulder girdle, and often the weaker arm.
- Dynamic strengthening of the neck and scapular muscles against light resistance (band, small weights): 3 × 10–12, 3 times a week, increasing the load when 12 repetitions become easy.
- Deep flexor endurance: 10 s holds, 3 × 8–10, daily.
- Upper limb strengthening (pushing, pulling) if a loss of strength persists: 3 × 8–12, 2–3 times a week.
Overall progression criterion: increase load or range only when the current level is pain-free and well controlled over 2–3 consecutive sessions.
One honest qualification: in that same meta-analysis, the gain in function measured by the Neck Disability Index reaches −3.60 points: statistically significant, but below the threshold of clinical relevance (MCID ≈ 5 to 7.5 points). The authors judge the level of evidence low and speak of a treatment that « may help ». The effect on pain is solid; the functional benefit on its own remains modest and uncertain. Hence the value of combining exercise with other techniques.
Should other techniques be added?
Often, yes, but we are talking about techniques, not about setting methods against each other. Manual therapy combined with exercise gives good results 5; in a recent network meta-analysis, manual therapy comes top of the probability ranking for pain and disability, ahead of exercise alone 6. A word on how to read this: these are probabilities of being the best treatment in the network, not a demonstrated superiority, and on only 8 studies. Conversely, adding mechanical cervical traction to a manual therapy + exercise programme brings no additional benefit in patients with cervical radiculopathy 7: no need to cling to it.
| Ingredient | What the evidence says |
|---|---|
| Progressive active exercise | Reduces pain, moderate to low evidence |
| Exercise + manual therapy / neural mobilisation | The best-supported combination |
| Cervical traction added | Benefit not demonstrated, uncertain |
Everyday positions and sleep
- Screen at eye level and regular breaks: avoid long hours with the neck bent forward.
- Sleep: a pillow that keeps the neck in line (neither too high nor too flat). On your back, a small support under the nape; on your side, the pillow fills the gap between shoulder and head. Some people relieve the arm by resting it on a cushion.
- During a flare-up, alternating positions and moving a little often beats prolonged stillness.
Returning to sport
The return is gradual and follows the same logic of increasing load: you reintroduce endurance and controlled movements first, then intensity, keeping to the acceptable-pain rule. Overhead activities and contact sports come back last, once strength and freedom from pain have returned. No movement becomes permanently « forbidden »: the aim is to reaccustom the tissues, not to protect them indefinitely.
The engine of treatment is not a miracle technique, but an active programme carried out patiently.
The common mistakes to avoid: waiting for the pain to disappear completely before moving; chasing the strong stretch that « pulls » in the arm (at the risk of irritating the root); persisting with a gadget supposed to « open » the foramen: adding specific techniques for that purpose has shown no additional benefit 5; and finally, stopping as soon as things improve: recovery is slow and strengthening continues after the pain has gone. And if surgery is raised, be aware that a structured physiotherapy programme gives, at two years, functional results comparable to those of decompression and fusion for cervical radiculopathy 3: all the more reason to give it its chance.
⚠️ The mistakes that keep the problem going
Cervical radiculopathy heals in the great majority of cases, but slowly: around 55 % of people report themselves recovered at 6 as at 12 months, with most of the progress happening in the first few months 1. That timescale is precisely what catches people out: when you expect a quick result, you multiply the missteps that make things drag. Here are the most common.
Passively waiting for it to « pass »
Resting completely and hoping it goes away on its own is the least effective strategy. In a landmark trial of 205 recent cases, both a semi-rigid collar with rest (3 to 6 weeks) and a physiotherapy programme with home exercises (6 weeks) reduced neck and arm pain markedly more than simply waiting 4. A period of taking it easy at the very start can be useful, but prolonged immobility and avoidance of movement keep stiffness and fear of moving going.
Looking for the « miracle » technique or machine
Many people hope for the single technique that will « put everything back in place ». The data invite modesty. In patients with cervical radiculopathy, adding mechanical cervical traction to a programme combining manual therapy and exercise brought no additional benefit for pain or function, at 2 as at 4 weeks 7. Likewise, adding manual techniques supposed to « open » the intervertebral foramen changed nothing further: what counted was the active programme itself, not any particular technique 5. Note: that trial compared two already active protocols, without a control group, and over the short term (weeks 4 and 8).
Believing that exercise « is no use »
That is the received idea to dismantle. A meta-analysis of 10 trials (871 patients) shows that exercise, alone or combined with conventional treatment, significantly reduces pain, with a large effect (SMD = −0.89) 2. Let us be honest about the limits: the measured gain in function (NDI −3.60 points) remains below the threshold of clinical relevance (around 5 to 7.5 points), the authors judge the level of evidence low and speak of a treatment that « may help ». Exercise is therefore no magic wand; but it does act on pain, and it need not be set against the other approaches.
Some reviews do place manual therapy top of the probabilities of being the most effective technique for pain and disability, ahead of exercise alone 6: careful, this is a probability ranking from a network of 8 studies, not a demonstrated superiority, and the authors find no significant difference between manual therapy alone and manual therapy with traction. The useful reading is not « this technique beats that one », but: active, multimodal management, combining exercises and techniques, is the most solid. That is also why a structured rehabilitation programme should precede any decision to operate on cervical radiculopathy 3.
Going too fast… or giving up too soon
Two mirror-image mistakes. Trying to « force » things to make up for lost time stirs up symptoms; conversely, stopping as soon as pain lessens deprives you of benefits that build over several months. The rule: progress in stages and keep going over time.
Key points
- Staying passive and waiting is the least effective strategy: moving, in measured doses, helps.
- No single technique or machine replaces an active programme; traction and specific techniques add nothing once that programme is in place.
- Exercise reduces pain and combines with the other techniques: they are not in opposition.
- Recovery is slow (≈55 % recovered at 6-12 months): patience and consistency are part of the treatment.
The signals that should make you ease off
Progressing does not mean ignoring your body. A few simple markers for slowing down and having your situation reassessed by a professional: arm pain that clearly and lastingly intensifies after sessions instead of settling, the appearance or worsening of muscle weakness, or a spread of the pins and needles and numbness. These do not mean all is lost, remember that the presence of paraesthesia is rather associated with better recovery 1, but they justify adjusting the programme rather than pressing on stubbornly. That is the whole role of support: assessing, dosing, progressing and spotting what should raise concern.
To understand the mechanisms, the causes and the diagnosis of this condition, see the full article on cervical radiculopathy.
🩺 Expected duration and the physiotherapist’s role
One question always comes back: « how long before this goes away? » The honest answer is: often several weeks to several months, with improvement that starts early but spreads over time. The good news is that under conservative treatment, without surgery, the prognosis of cervical radiculopathy is broadly favourable.
Realistic timescales, not magic ones
In a prognostic study 1, around 55 % of people said they were recovered at 6 months, and the proportion remained similar at 12 months: most of the progress happens in the first few months, then things settle slowly. Two useful markers emerge: symptoms that have dragged on before treatment are associated with poorer recovery, whereas the presence of pins and needles (paraesthesia) is rather a good sign.
In the early phase (symptoms of less than a month), a landmark trial of 205 patients 4 showed that a semi-rigid cervical collar with rest for 3 to 6 weeks, or physiotherapy with home exercises for 6 weeks, reduced pain markedly more than simply waiting: around –17 mm and –14 mm on the neck pain scale. In other words, acting early and staying active makes a measurable difference.
The right treatment is active, progressive and tailored to you, not a miracle technique.
What exercise brings, and its limits
A meta-analysis of 10 randomised trials 2 confirms that exercise, alone or combined with conventional treatment, significantly reduces pain, with a large effect size (SMD = –0.89). This is a solid and reassuring result: supervised movement brings pain down. On function, however, let us be transparent: the gain measured on the Neck Disability Index (–3.60 points) is statistically significant but remains below the threshold considered clinically relevant (around 5 to 7.5 points). The authors themselves judge the level of evidence low and speak of a treatment that « may help », probable help, not a demonstrated functional benefit.
A blinded trial 5 reports improvements in pain and limitations that are both statistically and clinically significant at 4 and 8 weeks. Two qualifications: both groups followed an active programme (no untreated group), and follow-up stopped at the short term. These figures therefore reflect the progress observed under an active programme, not a proven superiority over the natural course.
An active, multimodal approach
The best data do not pit « one technique against another »: they show approaches that combine. Exercise goes together with manual therapy and neural mobilisation. Take care in reading comparisons of techniques: in a 2025 network meta-analysis 6, manual therapy comes top of the probability ranking of being the best treatment for disability (68.1 %) and pain (59.5 %), ahead of exercise alone. These are rank probabilities, not a proven difference, and the authors note there is no significant difference between manual therapy alone and manual therapy + traction. The level of evidence, across 8 studies, remains limited.
| Approach (technique) | What the evidence says | Level |
|---|---|---|
| Manual therapy + exercise | Reduce pain and limitations in the short term 5 | Moderate |
| Exercise (alone or combined) | Reduces pain; functional gain below the clinically relevant threshold 2 | Low |
| Cervical traction added | No additional benefit demonstrated in patients with cervical radiculopathy 7 ; uncertain value 6 | Low |
| Structured physiotherapy vs surgery (ACDF) | Equivalent on function at 2 years 3 | Moderate |
Two practical lessons. First, adding intermittent cervical traction to a manual therapy + exercise programme brought no additional benefit for pain or disability in patients with cervical radiculopathy 7. Second, on measures of function (cervical range of motion, muscle endurance, hand function), a structured physiotherapy programme did as well as decompression and fusion surgery at 2 years 3 : the authors recommend trying a structured programme before considering a decision on ACDF for cervical radiculopathy. On pain, however, surgery may bring earlier relief in the first year.
The physiotherapist’s role
None of this comes down to exercises printed on a sheet of paper. A physiotherapist’s value is to assess your particular situation, to dose the load, to progress the programme at the right pace, to lift the fear of moving and to make you independent, while watching for warning signs. The exact content depends on their clinical reasoning in your case; that is precisely the value of follow-up rather than a fixed protocol. To understand the diagnosis, the mechanisms and the assessment, see the full article on cervical radiculopathy.
When to seek advice without delay
Most cases of cervical radiculopathy run a favourable course with an active programme. Some signs should nonetheless prompt prompt advice, without waiting for it to « sort itself out »: muscle weakness in the arm or hand that is setting in or worsening, problems affecting both limbs, difficulty walking or coordinating movement, problems passing urine, or sudden, unusual pain with fever or feeling generally unwell. In these situations, medical advice takes precedence over exercises.
Key points
- A favourable but slow prognosis: around 55 % of people recovered at 6 and 12 months, with most of the progress in the first few months 1.
- Exercise does reduce pain (SMD –0.89); the measured functional gain remains below the clinically relevant threshold, on low-level evidence 2.
- The winning approach is active and multimodal: exercise combines with manual therapy and neural mobilisation; added traction has shown no additional benefit 7.
- A structured physiotherapy programme deserves to be tried before considering surgery for cervical radiculopathy 3.
- Worsening weakness, both arms affected, problems walking or passing urine: seek advice without delay.
Bibliography
Every reference checked individually on PubMed (clickable PMID). 7 sources. Click a superscript note marker in the text: the bibliography opens and highlights the source.
- Sleijser-Koehorst MLS, Coppieters MW, Heymans MW, et al. (2018). European Spine Journal. PMID 30327908. doi:10.1007/s00586-018-5777-8.
- Liang L, Feng M, Cui X, Zhou S, Yin X, Wang X, Yang M, Liu C, Xie R, Zhu L, Yu J, Wei X (2019). Medicine (Baltimore). PMID 31702624.
- Peolsson A, Söderlund A, Engquist M, Lind B, Löfgren H, Vavruch L, Holtz A, et al. (2013). Spine (Phila Pa 1976). PMID 23407407.
- Kuijper B, Tans JT, Beelen A, Nollet F, de Visser M (2009). BMJ. PMID 19812130.
- Langevin P, Desmeules F, Lamothe M, Robitaille S, Roy JS (2015). Journal of Orthopaedic & Sports Physical Therapy. PMID 25420010.
- Xu X, Yan L (2025). Journal of Pain Research. PMID 40255362.
- Young IA, Michener LA, Cleland JA, Aguilera AJ, Snyder AR (2009). Physical Therapy. PMID 19465371.
❓ Frequently asked questions
Which exercises are recommended for cervical radiculopathy?
The data support active exercise programmes, alone or combined with manual therapy, including home exercises 4. A blinded trial shows that it is the structure of the active programme that counts, not any particular manual technique: adding techniques supposed to « open » the intervertebral foramen brings no additional benefit 5. The precise choice of exercises follows from an individual assessment by a physiotherapist.
Is exercise really effective against cervical radiculopathy?
Yes for pain: a meta-analysis of 10 randomised trials (871 patients) finds a significant reduction in pain with exercise (SMD = -0.89; 95 % CI: -1.34 to -0.44). The gain in function on the Neck Disability Index (-3.60 points) is statistically significant but remains below the threshold of clinical relevance (MCID ≈ 5 to 7.5 points), and the authors describe the level of evidence as low, concluding that exercise « may help » 2.
How long does it take to recover from cervical radiculopathy?
The prognosis under conservative treatment is broadly favourable but slow: around 55 % of patients report themselves recovered at 6 and at 12 months, with most of the improvement occurring in the first few months. A longer duration of symptoms increases the risk of poor recovery, whereas the presence of paraesthesia reduces it 1.
Is cervical traction useful for cervical radiculopathy?
Its value is not demonstrated. In 81 patients with cervical radiculopathy, adding intermittent mechanical cervical traction to a programme of manual therapy and exercise brings no additional benefit for pain, function or disability at 2 and 4 weeks 7. A 2025 network meta-analysis likewise concludes that the benefit of traction remains uncertain and requires more evidence 6.
Physiotherapy or surgery for cervical radiculopathy?
A prospective randomised study with 2 years of follow-up finds no significant difference between physiotherapy alone and anterior cervical decompression and fusion (ACDF) followed by physiotherapy on functional outcomes (cervical range of motion, muscle endurance, hand function; P = 0.17 to 0.91). The authors conclude that a structured physiotherapy programme should precede a decision on ACDF for cervical radiculopathy. This comparison is on function, not pain, and on a small sample 3.
Behind this article
An author who explains, a reviewer who checks.
How we write and check our content
Physiotherapist · co-founder of Physio Learning
Marked for life by his first four-hour lecture without a single image, he took a master’s in instructional design so that it would never happen to anyone again. He hunts down publication bias and unreadable slides with the same intransigence.
Head of scientific content
Physiotherapist specialising in neuro-musculoskeletal practice and holder of a master’s in public health. He checks the methodological rigour of every article: primary sources, levels of evidence, no exceptions.