The Neck You Never Programmed. The Only Structure Holding Up The Part Of You That Makes Every Decision.
Seven vertebrae, about a dozen muscle pairs and no slot in anybody's program. The neck carries a four to five kilogram head, holds your eyes level while the rest of you moves, and houses one of the densest fields of position sensors in the body. This article explains what the structure is, what the research says about training it directly, and gives you a protocol that takes six minutes twice a week.
Opening. The seventieth minute
Marc is thirty one and plays rugby union at a competitive club grade. He squats and pulls numbers most of his team cannot, he has a programmed slot for his rotator cuff, he does his ankle work, and he has read enough to know why all of it is there.
In the seventieth minute of a hard game his head starts to drop.
Not dramatically. His chin creeps forward, his gaze lowers a few degrees, and he enters contact a little lower and later than he did in the twentieth minute. On Monday his neck is stiff and he files it as part of playing, which it is, and as nothing to act on, which it is not.
Nobody has ever written him a neck session. He has a program covering twenty muscle groups and the structure holding up his head is not one of them. That is close to universal, and the neck is the only load bearing region most serious lifters wait to be told about by pain.
Section one. What the structure actually is
Seven cervical vertebrae, carrying a head of roughly four to five kilograms in an adult.
The top two are unlike anything else in the spine. The atlas sits under the skull and is built for nodding. The axis beneath it carries a vertical peg the atlas rotates around, and that single joint provides a large share of your total head rotation. The five below share the rest, along with flexion, extension and side bending, which is why neck range is a distributed property rather than a single hinge.
Around that column sit two very different muscle layers, and the distinction is the most useful thing in this article.
The superficial layer is the one you can see. Sternocleidomastoid at the front, the scalenes at the side, upper trapezius and the long extensors behind. These are the big movers, and the ones that take over when something is missing.
The deep layer sits directly against the vertebrae. Longus capitis and longus colli at the front, the small segmental extensors and the suboccipital group behind. These are short, they cross one or two segments, and their job is not movement. It is holding the column in order while the big muscles work and the rest of the body moves underneath.
One layer generates. One layer governs. Almost all the interesting research is about the second one.
Section two. The specificity finding that should have changed programming
Here is the finding that disposes of the most common excuse for skipping neck work.
Most lifters assume the neck is covered by what they already do. Heavy shrugs, upright rows, farmer carries, bracing under a squat bar. Something in that must be loading it.
Conley and colleagues tested exactly that and published the answer in 1997, across a pair of papers in the European Journal of Applied Physiology and the Journal of Applied Physiology. They compared a resistance program including shrugs and upright rows against one that added direct neck extension work. The group performing direct neck work gained neck extensor size and strength. The group doing shoulder and upper back work without direct neck loading did not gain in the neck muscles to the same degree.
The principle is ordinary and the application is almost universally ignored. A shrug loads the shoulder girdle. It does not load the cervical extensors the way moving the head against resistance does.
So if your neck has never had a set written for it, the honest position is that it has never been trained, regardless of what else is in the program.
Section three. The layer that goes quiet
Now the clinical literature, which is where the deep layer earns its reputation.
Gwendolen Jull and colleagues at the University of Queensland built the standard assessment, the craniocervical flexion test. It asks a person lying on their back to perform a small controlled nod, a deep flexor action, without recruiting the big superficial muscles. The repeated finding across that body of work, with Deborah Falla a central contributor, is that people with persistent neck pain show reduced capacity in the deep cervical flexors and greater superficial muscle activity during the same task.
Read the direction carefully, because the research is careful about it. Whether the deep layer going quiet causes the pain or follows it is not settled, and both are probably true in different people. The pattern itself is real, measurable and responsive to specific training.
This is also the mechanism under the sustained position problem. Hold your head still for hours, which is what a desk, a windscreen and a phone all require, and the deep layer is asked to do a job it is not conditioned for while the superficial layer takes over. We wrote the broader argument in The Posture Decade.
One note on the figures that circulate here. You will have seen claims that a head held forward of the shoulders loads the neck with a specific multiple of its weight. The direction is sound physics. The numbers trace back to a single widely criticised paper, so take the principle and leave them alone.
Section four. The neck as a sensor
The part almost nobody knows, and the reason this is not just a strength topic.
The deep muscles of the neck, particularly the suboccipital group under the skull, carry an extraordinary density of muscle spindles. Spindles report muscle length and rate of change, and per gram of tissue this region is among the richest in the body.
That makes anatomical sense. Your inner ear reports head motion, your eyes report the world, and your neck reports where your head sits relative to your body. Those three streams are compared continuously, and balance and gaze stability depend on them agreeing.
When neck input is disturbed, the agreement degrades. Julia Treleaven, also at the University of Queensland, has published extensively on altered cervical proprioception and its relationship with dizziness, unsteadiness and disturbed gaze control in people with neck disorders. A recognised clinical picture rather than a fringe idea.
For an athlete the reading is this. A fatigued neck is not only a weaker neck. It is a less accurate sensor feeding the system that keeps your eyes level, which is why Marc's seventieth minute is a gaze problem as much as a strength problem.
Section five. The collision question, handled carefully
This needs to be stated precisely, because it is the claim most often overstated.
There is observational evidence associating greater neck strength with a lower reported rate of concussion. The study most often cited is Collins and colleagues, published in the Journal of Primary Prevention in 2014, which looked at high school athletes across several sports and found a lower reported incidence among those with greater neck strength. Christopher Hrysomallis reviewed the broader question in Sports Medicine in 2016 and described the rationale as plausible and the evidence as limited.
The rationale is straightforward. A stiffer head and neck segment behaves as a larger effective mass, so a given force produces less head acceleration.
Now the boundaries, and they matter.
That is association in observational data, not a randomised trial, and nobody has run the trial that would settle it.
Neck training does not stop concussion happening and nothing here should be read that way. Concussion is a medical matter, managed by return to play protocols set by your code and by clinicians.
And more neck size is not automatically more protection. The mechanism depends on the segment being stiff at the moment of contact, which is a timing and awareness question as much as a strength one.
The defensible statement is narrow and still worth acting on. Neck capacity is trainable, it is associated with better head kinematics under load, and six minutes twice a week costs nothing.
Section six. A defensible protocol
Two slots a week, six minutes each, attached to sessions you already run. Start with isometrics, the safest entry and the easiest to control.
Begin with the deep layer. Lie on your back and perform a slow small nod, as though saying yes to someone standing above you, without lifting your head off the floor. Hold ten seconds, six to eight holds. The sensation should be deep and at the front of the throat, not in the big muscle at the side. If the superficial muscle takes over, you have gone too far into the range.
Then load the four directions. Self applied pressure from your own hand, held for ten to fifteen seconds in flexion, extension and both sides. Two rounds. Build pressure gradually, and never load a direction into pain.
Progress to slow movement at about week four. Controlled range against light resistance, three sets of eight to twelve, slow in both directions, in flexion, extension and rotation. A head harness, or a padded plate on the forehead, is sufficient. Increments here are smaller than anywhere else in your program, and the rotation matters most, because the top two segments do most of the turning and they are what sedentary life takes away first.
Where it sits. After upper body sessions, never before.
What it is not. This is not care for a painful neck. Numbness, pins and needles, radiating arm symptoms, dizziness or headache after a session go to a physiotherapist before you load anything. The deep flexor work in particular is something a clinician can calibrate in one appointment and you can get subtly wrong for months.
The inputs are the same as everywhere else. Adequate daily protein and creatine monohydrate to underwrite the work. Nothing is specific to the neck, and any product claiming to be should be read with that in mind.
Section seven. The mastermind frame
There is a pattern in how lifters choose what to train, and it is not based on importance.
We train what we can see, measure and compare. The chest has a bench press and a number. The back has a row and a number. Grip turned out to be one of the better documented markers of how a body is ageing, which is why it earned its own piece. The neck has none of that. No competition lift, no mirror angle, no leaderboard, and therefore no slot.
So the structure holding up your skull is governed entirely by whether anybody ever told you to work on it. For most people that arrives in their forties, from a physiotherapist, after a year of something that would have been easier to handle at thirty.
Marc does not need a neck like a prop forward. He needs six minutes, twice a week, for twenty years, in a region that currently gets none.
Train the thing holding up the part of you that makes every decision. Nobody will notice, which is the usual signature of work that matters.
Frequently asked questions
Will direct neck training make my neck look thick?
Over months of consistent loading some visible change in the extensors is likely, because it is muscle and it responds to load like muscle. The protocol here is a capacity and control programme rather than a size programme, and the time involved is far too small to produce the look most people worry about.
Is a neck harness necessary?
No. Self applied isometric pressure costs nothing and covers the first several weeks properly. A harness or a padded plate is simply a convenient way to add small controlled increments once isometrics are easy, and plenty of people never need one.
I have a stiff neck most mornings. Should I start this?
Not on your own. Persistent stiffness, pain, headache or any arm symptoms deserve an assessment first, because the right starting point for an irritable neck is different from the right starting point for a quiet one. A physiotherapist can sort that out quickly and will usually give you a version of the deep flexor work anyway.
Does neck strength help my squat or deadlift?
There is no good evidence that it improves either lift, and we are not going to pretend otherwise. Head position does influence how a heavy lift feels and tracks, and a neck that tolerates sustained bracing is a comfort advantage over a long block. File it as capacity and durability, not as performance.
How long before anything changes?
Control and endurance in the deep layer respond quickly, often inside four to six weeks. Measurable size and strength in the extensors follow the same slow timeline as any other muscle, which means months rather than weeks. Nothing about the neck is faster than the rest of you.
Written for the Supplement Superstore Vault. We sell the supplements that support the work. We do not sell the work. Information here is educational and is not medical advice. Speak with a qualified health professional before changing any protocol, especially during pregnancy, breastfeeding, competitive training or while managing any clinical condition.
