The multifidus is the muscle that over the last twenty years has taken on the role of favourite culprit of back pain: deep, invisible, easy to picture as the hidden gear that has jammed. Entire exercise protocols have been built on this idea, each with the promise of "waking it up".
It is worth saying this straight away, because it changes how you will use the rest of the article: that promise does not hold up as it is told. The link between the multifidus and low back pain exists but is weaker and more nuanced than what is read around, and above all it has not been shown that the benefit of exercise comes from waking that muscle up. This does not mean that the exercises you will find below are useless: it means they work for a different reason from the one you were told, and knowing which one keeps you from losing months chasing a sensation.
⚠ Important: read this before continuing
The exercises and advice in this article are general guidance and may not suit your situation. Every person has a unique clinical history: what works for others could be counterproductive for you.
Want a professional's opinion? Start with the questionnaire: you sign up and see who will follow you, a physiotherapist registered with the Albo chosen for your case, whom you can write to in chat. If you would rather hear a voice first, the first consultation is free and we look at your situation together.
First of all: the signs that should not be managed with exercises
Back pain is almost always benign. The Lancet series on low back pain says so without beating about the bush: in the vast majority of people it is not possible to identify a specific nociceptive cause, and only a small share has a well-understood pathological cause, such as a vertebral fracture, a tumour or an infection. It is precisely because that share is small that it needs to be recognised straight away, before you get down on the floor to do exercises.
See a doctor, and not a strengthening programme, if even one of these pictures appears:
- Loss of strength in the leg or foot that worsens over days, for example difficulty lifting the tip of the foot or climbing stairs.
- Numbness in the area in contact with the saddle (inner thighs, perineum), or difficulty controlling urine or stools: this is an emergency, go to the emergency department.
- Pain that appeared after a major trauma, a fall or an accident, especially if you have osteoporosis or have taken corticosteroids for a long time.
- Constant night pain that does not change in any position, particularly if accompanied by fever, general malaise or unintended weight loss.
- A history of cancer, even a distant one.
There is also a less dramatic but equally useful criterion. The review of international guidelines for primary care, which compared fifteen of them, recommends looking for warning signs with history and physical examination, not using routine imaging, assessing psychosocial factors, and referring to a specialist if there is no improvement after four weeks. Four weeks without change is not your failure: it is information, and it should be used to revise the starting hypothesis.
What the multifidus is and what it really does
The multifidus runs along almost the whole spine, from the sacrum up to the upper cervical vertebrae: it stops at the axis, the second cervical vertebra, and does not reach the skull. It is organised in short bundles that span only a few vertebrae at a time. It is one of the deepest muscles of the back, among those closest to the joints.
This architecture explains its function. Unlike the superficial muscles, it does not produce large movements: its lever arm is too short. It works instead on the fine control of micro-movements between one segment and the next, and on segmental extension. In the lumbar region its bundles are particularly bulky, and this is why it has become the protagonist of the discussion on spinal stability.
Beware, however, of a logical step that is taken too quickly. "It contributes to stability" does not mean "it is responsible for stability". The spine is stabilised by a system that involves all the trunk muscles, abdominal pressure, the thoracolumbar fascia and the nervous control that coordinates them. The multifidus is an important piece of that system, not the switch that turns it on.
The awkward question: does a weak multifidus cause back pain?
Here we need to look at the numbers, because they are more interesting than the slogan.
A meta-analysis published in The Spine Journal in 2022 collected imaging studies comparing the paraspinal muscles of people with low back pain with those of people without it: twenty-five studies included, twenty ended up in the meta-analysis. The result for the total cross-sectional area of the multifidus is an average standardised difference of -0.24, with a 95% confidence interval from -0.5 to 0.03. The interval includes zero: the muscle tends to be a little smaller, but the figure does not reach statistical significance. The solid result, instead, concerns the fat inside the muscle, with an effect of medium size (SMD 0.61; 95% CI 0.30 to 0.91). For the erector spinae and the psoas no difference linked to low back pain emerged.
A systematic review from 2016 in Pain Physician adds a detail that almost nobody reports, and it is what dismantles the most widespread premise. Distinguishing the types of low back pain, it found moderate evidence of multifidus atrophy in chronic low back pain, but equally moderate evidence that in recurrent and acute low back pain atrophy is not seen in any lumbar muscle. If your back pain comes and goes in episodes, which is the most common case, the picture of "your multifidus has atrophied" is not what the literature describes. The part that does not favour this reading must be said too: in recurrent low back pain the same review reports a higher index of muscle fat in the erector spinae, multifidus and paraspinal muscles, that is, a relative share of lipids inside the larger muscle, even without true fatty infiltration. Something in the muscle changes; what does not hold up is the story of atrophy.
And there is a limitation that applies to both reviews: the included studies are case-control, snapshots taken of people who already have pain. They capture a difference; they do not say whether the muscle changed before the pain, or whether it is the pain, with the resulting reduction in activity, protective guarding and time spent still, that changed the muscle. The second hypothesis is at least as plausible as the first, but the first sells much better.
An important point: before trying these exercises, make sure you have received a correct diagnosis. People often assume they have one problem, but the real cause of the pain is something else. Doing the wrong exercises can make things worse instead of better.
Why "waking up the multifidus" is not the mechanism we thought
Let us suppose, for a moment, that the multifidus is involved. What remains to be shown is the next step: that exercise works because it changes that muscle. This is the part of the theory that has been tested directly, and it is the part that has held up least.
A systematic review from 2014 in the Journal of Pain selected fifteen studies to answer exactly this question: when a conservative treatment changes the transversus abdominis or the lumbar multifidus, does that change explain the clinical improvement? The conclusion is clear for the transversus: strong evidence that variations in thickness during contraction and in anticipatory activation are not related to changes in pain and disability, and it remains uncertain for the multifidus, where the results on morphometry are conflicting. In short: the muscle can change too, but it is not that change that brings the improvement you hope for.
The second blow comes from the treatments themselves. The 2016 Cochrane review on motor control exercise, the family to which the multifidus exercises belong, those that start from activating the deep trunk muscles and then integrate them into more complex tasks, included 29 randomised trials and 2,431 patients with chronic non-specific low back pain. The results, in order:
- compared with other forms of exercise, they are not clinically more effective, at any follow-up point and on any measured outcome (evidence of low to moderate quality in the authors' conclusions);
- compared with minimal intervention, they are effective on pain in the short, medium and long term, with effects of medium size (in the long term, mean difference -12.97; 95% CI -18.51 to -7.42), and a clinically important benefit also on function and perceived global recovery, although here the quality of the evidence is only low to moderate;
- compared with manual therapy, evidence from moderate to high that there is no clinically important difference;
- compared with exercise plus instrumental physical therapies, motor control, by contrast, is superior on pain, disability, perceived recovery and quality of life, with medium to large effects (short-term pain, mean difference -30.18; 95% CI -35.32 to -25.05): this is the comparison that goes best for it, but it is also the one with the most fragile evidence, of very low to low quality;
- the adverse events reported in the studies were minimal or absent.
The authors' conclusion is the sentence that should sit at the top of every article like this one: since motor control is not superior to other forms of exercise, the choice should depend on the preferences of the patient and the therapist, the training of the person leading it, costs and safety.
The opposite must also be said, because it exists. A critical synthesis published in the Journal of Athletic Training in 2017, which analysed a 2012 meta-analysis of five studies and 414 patients, found an advantage of core stability exercise over general exercise at three months, both on pain (mean difference -1.29; 95% CI -2.47 to -0.11) and on function (-7.14; 95% CI -11.64 to -2.65). But the same work shows that the advantage fades: already at six months the difference is no longer significant either on pain (-0.50; 95% CI -1.36 to 0.35) or on function, and it is not significant at twelve months either (-0.32; 95% CI -0.87 to 0.23). Two caveats before taking this figure at face value: the authors point out that all the included studies were rated at high risk of bias, and the summary is itself inconsistent, because at one point it reports values at six and twelve months and at another it states that function was not measured at those time points. An initial advantage that disappears is still useful information: it may be worth starting here, but there is no sense in staying here forever.
So do these exercises help? Yes, for a different reason
If you remove the theory of the muscle as a switch, one very solid thing remains: for chronic low back pain, exercise is recommended by guidelines, and it is among the few measures with a favourable benefit-risk ratio.
A network meta-analysis from 2023 of 75 randomised trials and 5,254 participants compared twenty different types of exercise. Core and stabilisation exercises improved physical function more than conventional rehabilitation (SMD -1.04; 95% CI -1.80 to -0.28) and more than doing nothing (SMD -1.13; 95% CI -1.66 to -0.59). On pain, however, the only ones to outperform conventional rehabilitation with a statistically significant result were tai chi, yoga, Pilates and sling exercise, although in the conclusions the authors then list among the approaches that improve chronic low back pain also motor control and core and stabilisation exercises, in a wording more generous than their own comparisons show. The same authors urge caution because of the quality and quantity of the included studies, but the underlying message is consistent with the Cochrane review: what matters is moving progressively, much more than which label the protocol carries.
That is why it makes sense to keep reading. The exercises for the multifidus that you will find below are a good starting point not because they hit some secret target, but because they are low-load, easy to learn without equipment, not threatening for people who are afraid to move, and can be arranged in a clear progression. If someone has prescribed a multifidus programme for you, in short, it is not a wrong programme: it is simply a programme whose value lies somewhere other than how it was presented to you. They are a way to get a back that has stopped moving going again. That is already a lot.
Feeling the area, without the illusion of isolating a muscle
The classic gesture is this: lie on your back with your knees bent, place your fingertips at the sides of the lumbar spinous processes, in the groove between the spine and the muscle mass. Try to imagine lengthening the spine and "bringing together" two vertebrae, without moving the pelvis, without pushing your feet into the floor, without holding your breath. If you feel a slight firmness under your fingers, that is the area you are engaging.
To be honest: that firmness is not proof that you are activating the multifidus and nothing else. Under your fingers there is a layer of overlapping muscles, and the multifidus is among the deepest. In studies activation is measured with real-time ultrasound or electromyography, not with the fingertips, and, as we have seen, even when it is measured properly its change does not reliably follow clinical improvement. Use palpation for what it is really worth: a way to slow down, breathe and not recruit your whole body into an exercise that should be subtle.
A last note on language. Cues such as "draw the navel towards the spine" serve to reduce global stiffening, not to select a muscle. If the cue helps you move better, keep it; if it leads you to hold your breath and stiffen up, drop it without any guilt.
Six exercises, in progression
These are the multifidus exercises most used in practice, ordered from the simplest to the most demanding. The repetition numbers are practical working conventions, not proven dosages: no study has established the optimal dose. Use them as a starting point and adjust them to your response, bearing in mind that mild discomfort that fades within a few hours is acceptable, while pain that increases from day to day tells you that you have gone too fast.
1. Supine activation. Lying on your back, knees bent, feet on the floor. Find the light contraction described above and maintain it while you breathe normally for about ten seconds, then let go completely. Repeat ten times. The goal is not strength, it is learning to switch on and off.
2. Bridge. From the same position, lift the pelvis, resting on your heels, until shoulders, pelvis and knees are aligned, then lower down vertebra by vertebra. Ten to fifteen slow repetitions. It is the first exercise in which the back has to stay orderly while something really moves.
3. Bird dog. On hands and knees, extend one arm and the opposite leg while staying low, without arching or rotating the pelvis. Ten repetitions per side. If you want to work on it well, the guide dedicated to the bird dog goes into detail on the most common mistakes.
4. Dead bug. Lying on your back, arms towards the ceiling and hips and knees at 90 degrees. Slowly extend one arm and the opposite leg without the lumbar area lifting off the floor, then return. Ten repetitions per side. Here too there is a deep dive on the dead bug if you want to refine the execution.
5. Plank. Resting on your forearms and the tips of your toes, keep one straight line from shoulders to heels, without letting the pelvis drop. Thirty to sixty seconds, or shorter sets repeated if the position makes you lose alignment.
6. Pallof press. With a band anchored to the side at chest height, hold your hands at the centre of your sternum and push forward, resisting the rotation the band imposes on you. Ten to twelve repetitions per side. It is the only exercise on the list that introduces an external load, and it serves to bring control into a more realistic situation.
These are six exercises, and in this sequence lies the one thing the literature supports clearly: a progression. You start with little, you add movement, then you add load.
How to put a programme together
Three or four sessions a week is a sustainable rhythm for most people. Here too, as with the repetition numbers, this is a practical convention and not a frequency shown by a study. Consistency matters more than the length of a single workout. Stay on the first two or three exercises until you perform them without obvious compensations, then add the next ones.
Two practical pointers are worth keeping in mind. The first: do not turn everything into a hunt for the perfect sensation: if every repetition becomes an activation test, the exercise stops being movement and becomes anxiety. The second: always combine it with something you enjoy, such as walking, swimming, yoga or cycling. The 2023 network meta-analysis shows precisely that several routes lead to the same result, and the best route is the one you keep coming back to.
If you have a precise diagnosis behind you, the context changes. After a disc herniation, or in a picture of chronic low back pain, progression must be built on the symptoms you have, not on a generic table.
Bringing it into everyday life
The part that really pays off is not the multifidus exercise session: it is what happens in the other twenty-three hours. There is no need to "activate the multifidus" before every movement. It is an instruction that not only has no solid basis, but tends to make people stiffer and more worried about their body.
What helps instead is variety: getting up often if you work sitting down, changing position before your back starts to complain, lifting weights close to the body while hinging at the hips, walking regularly. And there is one last thing, which the Lancet series highlights: high initial pain intensity, psychological distress and pain in several parts of the body increase the risk that a low back pain becomes persistent and disabling. Sleep, stress and fear of movement are not background noise; they are part of the problem, and they often weigh more than the cross-sectional area of a deep muscle.
When an assessment is needed
Beyond the signs listed at the beginning, a physiotherapy assessment does the thing that a list of exercises cannot do: it works out which movements reproduce your pain, which ease it, what in your day keeps the problem going, and from there builds a load progression that makes sense for you. Much of this work can also be done by video call, because it rests on observing movement and correcting execution; when manual examination is needed, however, an in-person visit remains the right route.
In short
The multifidus is a real muscle, with a real function, and it is reasonable to include it in a back programme. But the story of the sleeping muscle to be woken up finds no confirmation: the difference compared with people without pain is modest and concerns mainly chronic low back pain, the studies do not say what came first, and changes in the muscle do not reliably explain clinical improvement. Motor control exercises beat doing nothing and are on a par with manual therapy and with other forms of exercise.
The right way to use this page, then, is not to look for the perfect activation: it is to take the multifidus exercises as a safe progression to restart from, and then add the movement you are willing to keep up over time. If nothing changes after four weeks, the right step is not to push harder: it is to get checked again.
You might also like:
Pre-Workout Stretching: What to Do and What to Avoid Home Exercises: A Complete Programme with No Equipment Swimmer's Shoulder: Prevention, Symptoms and How to Treat It
Stay in touch with us Join our dedicated Facebook group: we answer all patients, give practical advice and clear up the most common doubts.
Sources
- Alteration of lumbar muscle morphology and composition in relation to low back pain: a systematic review and meta-analysis, The Spine Journal, 2022
- Structural Changes of Lumbar Muscles in Non-specific Low Back Pain: A Systematic Review, Pain Physician, 2016
- Do changes in transversus abdominis and lumbar multifidus during conservative treatment explain changes in clinical outcomes related to nonspecific low back pain? A systematic review, The Journal of Pain, 2014
- Motor control exercise for chronic non-specific low-back pain, Cochrane Database of Systematic Reviews, 2016
- Core Stability Exercise Versus General Exercise for Chronic Low Back Pain, Journal of Athletic Training, 2017
- Exercise intervention for patients with chronic low back pain: a systematic review and network meta-analysis, Frontiers in Public Health, 2023
- Clinical practice guidelines for the management of non-specific low back pain in primary care: an updated overview, European Spine Journal, 2018
- What low back pain is and why we need to pay attention, The Lancet, 2018
Last source check: August 2026.
Wondering how much it would cost to be followed? The price depends on how many sessions are needed, not on the single rate: the figures are in How much physiotherapy for back pain costs.
Need a consultation?
Start the questionnaire on Ri-Hub: a few questions about your case, you sign up, and you see who will follow you: a physiotherapist registered with the Albo, whom you can write to in chat whenever you like. If you would rather talk it through by voice, the first consultation is free, lasts twenty minutes and commits you to nothing: they arrive having already read what you wrote, and together you decide whether and how to go on.
