The McKenzie method: what the extension exercises are and who they suit

Search “exercises for back pain” and the McKenzie method’s signature move, a slow push-up from lying face down, appears within the first few results almost without fail. It has become one of the most widely recognised images in back pain self-treatment, shared and demonstrated far more often than the assessment process it actually comes from, which has left a lot of people trying the exercise without really knowing whether it suits their particular pain.

That gap matters, because the McKenzie method isn’t actually a specific exercise; it’s a system for working out which movements help a specific patient’s pain and which ones make it worse, with extension being only one of several possible answers that assessment can produce. Understanding the difference between the method and its most famous exercise explains both why it helps some people dramatically and why it does nothing, or occasionally worsens things, for others, and why two people with what looks like the same back pain can be given completely different exercise directions by the same clinician.

What the McKenzie method actually is

Developed by New Zealand physiotherapist Robin McKenzie after he noticed that a patient’s severe back pain improved unexpectedly after lying in a position of spinal extension, the method that bears his name is formally called Mechanical Diagnosis and Therapy. Its core idea is that spinal pain can often be classified by how it responds to repeated movements and sustained positions, and that this response, rather than an imaging finding alone, should guide which direction of exercise is prescribed.

A McKenzie-trained clinician’s assessment involves testing a patient’s pain response to a series of specific movements, usually starting with repeated spinal extension, and watching closely for two key signs: whether pain centralises, meaning it moves from a more distant location like the leg toward the spine itself, or peripheralises, spreading further away from the spine. Centralisation with a particular movement direction is generally a strong, reassuring sign that this direction will help; peripheralisation suggests the opposite direction should be tried, or that the pain doesn’t fit a pattern the method addresses well at all.

This assessment-first structure is the actual method. The extension exercise that most people associate with the name McKenzie is simply the most common outcome of that assessment, because disc-related lower back and leg pain, one of the most frequent presentations in general practice and physiotherapy, very often responds well to extension. But the same assessment process, applied to a different patient, can just as validly conclude that flexion, or a side-glide movement, is the direction that actually helps.

Who the extension exercises genuinely help

The classic McKenzie extension pattern, pain that improves or centralises with repeated extension and worsens with prolonged sitting or forward bending, is most commonly seen in patients with a disc bulge or early disc herniation where the disc material is still positioned in a way that extension can help manage. These patients frequently describe their pain as worse first thing after getting out of bed or after a long car journey seated, and better after standing and gently arching the lower back.

For this subgroup, the prone press-up, lying face down and slowly pushing the upper body up with the arms while letting the hips stay on the floor, performed gently and without forcing the range of motion, in sets of repetitions multiple times a day, can produce genuinely striking improvement within days, sometimes reducing leg pain to nothing while the lower back pain that remains is considered a better outcome because it’s centralised rather than peripheral. This pattern of rapid improvement is part of why the exercise has such a strong reputation among people who’ve tried it and had it work.

Patients with facet joint pain that’s aggravated by extension, with spinal stenosis, where extension narrows the already-tight spinal canal further, or with a disc pattern that actually centralises with flexion rather than extension, don’t fit this profile and can genuinely worsen with the same exercise that helps the disc-related subgroup. This is the central reason the assessment matters more than memorising which exercise is “for back pain.”

Why the assessment, not the exercise, is the real treatment

A trained McKenzie clinician doesn’t start a new patient on extension exercises automatically; they test the response to extension, flexion and side movements in the assessment itself, often within the first session, and base the prescribed direction on what the patient’s own pain pattern actually shows rather than a general rule about what usually works for most people. This individualised response is what separates a proper McKenzie assessment from simply recommending back extensions to anyone with lower back pain.

The speed of response during assessment is itself diagnostically useful. A patient whose leg pain genuinely centralises within a few repetitions of a test movement, becoming more localised to the lower back and less present in the leg, is giving the clinician useful real-time information about which direction of ongoing exercise is likely to help, in a way that an MRI scan, which shows structure but not necessarily what movement actually helps that structure feel better, often can’t provide on its own.

This is also why the method resists being reduced to a YouTube video or a printed handout alone. The handout can show the exercise; it can’t watch how an individual patient’s specific pain responds to it in real time, which is the actual clinical judgement the method depends on, particularly for patients whose pattern isn’t the straightforward, textbook disc presentation.

What it feels like to do the exercises correctly

For patients who are a good fit for extension, the prone press-up is typically done in sets of ten or so repetitions, several times through the day, with attention paid specifically to whether pain centralises or at least doesn’t worsen with each set. A mild increase in localised lower back discomfort during the exercise itself is often considered acceptable, even expected, as long as any leg pain is improving or staying the same rather than worsening or spreading.

This distinction, between a tolerable, centralising increase in local back discomfort and a worsening, peripheralising spread of pain, is the practical rule patients are taught to self-monitor by once a clinician has confirmed the direction suits them, and it’s a genuinely different skill from simply gauging whether an exercise “feels fine” in the moment. Stopping, or switching direction, the moment pain spreads further than before, rather than pushing through on the assumption that more repetitions will help faster, is part of how the method is meant to be self-applied safely between clinical sessions.

Postural advice usually accompanies the exercise prescription: avoiding prolonged slumped sitting, which tends to aggravate the same disc-related pattern the extension exercise is treating, and building brief extension movements into the day, standing and gently arching the back every so often, rather than relying on formal exercise sessions alone to counteract hours of flexed sitting. A lumbar roll or small cushion at the small of the back during long periods of sitting is a simple, commonly recommended way to maintain a small amount of extension throughout the day between formal exercise sessions.

What the evidence actually shows

Mechanical Diagnosis and Therapy has been studied for decades, and the evidence is reasonably consistent in a specific way: for the subgroup of patients who show a clear directional preference, meaning their pain reliably centralises with one direction and peripheralises with the other, outcomes tend to be good, often comparable to or better than generic exercise programmes. For patients who don’t show a clear directional preference during assessment, the evidence for forcing a particular direction is much weaker, which is consistent with the method’s own internal logic that not every patient fits a directional pattern at all.

This is a meaningfully different claim from saying “extension exercises cure back pain,” which is the version of the method that circulates most widely online. The actual evidence supports a more conditional statement: for patients whose assessment shows centralisation with extension specifically, extension-based exercise is a genuinely effective, well-supported treatment. For patients without that specific response, the evidence doesn’t support assuming the same exercise will help, and some research suggests it can aggravate symptoms in certain conditions.

Comparative studies against other forms of physiotherapy and generic exercise have generally found the McKenzie approach performs at least as well, and the directional-preference subgroup tends to improve faster, which is the practical argument for getting a proper assessment rather than skipping straight to a generic back exercise programme or, at the other extreme, assuming extension is the universal answer. Clinical guidelines in several countries now reference directional-preference testing as a reasonable part of initial assessment for mechanical lower back pain, alongside, rather than instead of, other established approaches.

How this fits into a broader treatment plan

For most patients with a clear directional preference, McKenzie-style exercise is often used as a first-line approach alongside general advice to stay active, avoid prolonged bed rest, and gradually return to normal activity, rather than as a replacement for every other aspect of back pain management. It works well specifically because it gives a patient an active, self-directed way to influence their own symptoms between clinical appointments, which tends to support confidence and reduce fear around movement, itself a meaningful part of recovering from mechanical back pain.

When a clear directional preference isn’t found, or when exercise alone doesn’t resolve symptoms within a reasonable timeframe, typically a few weeks of consistent, correctly performed exercise, the next step usually involves broader physiotherapy assessment, imaging if red-flag features are present, or referral to a spine specialist rather than persisting indefinitely with an approach that assessment has already suggested isn’t the right fit for that particular patient’s pain.

It’s also worth noting that a patient’s directional preference isn’t necessarily fixed forever. Someone who responded well to extension during an acute episode of disc-related pain may, months or years later with a different presentation, show a different or no clear directional response, which is why a fresh assessment at the start of a new episode is more reliable than assuming an exercise that worked once will always be the right one.

When extension is the wrong direction, or the wrong method entirely

Spinal stenosis is the clearest example of a condition where extension typically worsens rather than helps symptoms, because extending the spine narrows the already-reduced space available for the nerves passing through the spinal canal, the opposite mechanical effect from what helps a bulging disc. Patients with stenosis-related leg pain generally find relief in flexion, leaning forward, sitting, or the “shopping cart” posture of leaning on a trolley, rather than in extension, which is a pattern a proper assessment identifies quickly.

Pain that doesn’t fit cleanly into either a flexion- or extension-biased pattern, pain that doesn’t change meaningfully with any repeated movement, or pain accompanied by red-flag features like unexplained weight loss, fever, or progressive weakness, isn’t well suited to the McKenzie approach at all and needs a broader medical assessment rather than persisting with a mechanical exercise programme regardless of direction. The method is a tool for a specific, common category of mechanical spinal pain, not a universal answer to every back complaint.

For anyone trying extension exercises found online without a prior assessment, the safest approach is a cautious trial of a few repetitions, stopping immediately if pain worsens or spreads further down the leg, and seeking a physiotherapist’s assessment if pain hasn’t clearly improved within a few days. A proper assessment takes a single session and removes the guesswork of trying to self-diagnose which direction, if any, actually suits an individual case.

The method’s enduring popularity, decades after it was first developed, comes down to this combination of simplicity and precision: a small number of movements to test, a clear rule for interpreting the response, and an exercise programme a patient can largely manage themselves once the right direction has been identified. That combination is genuinely useful for the right patient, which is exactly why it’s worth confirming who the right patient actually is before assuming the exercise everyone has seen online applies to a particular case.