It usually starts somewhere around three in the afternoon. A tightness at the base of the skull, both sides, that creeps up over the back of the head and settles behind the eyes. It doesn’t stop you working. It just makes the last two hours feel like wading.
By evening it has gone, or a couple of paracetamol have seen it off. Then it comes back the next day at roughly the same time, and after a few months of that you start wondering whether the way you sit has something to do with it.
Can bad posture cause headaches? Yes. The mechanism connecting the two is well described, and the association turns up in study after study. But the research supporting “your posture is the problem” is thinner than most articles on this let on, and the research on what makes these headaches go away is unusually strong. Those two facts point in slightly different directions, which is the interesting part.
Below: what your neck is doing to your head, the two headache types that desk work drives, what the evidence really shows, how to work out whether posture is your culprit, and the fix that has the best trial evidence behind it.
The Short Answer
Poor posture is a recognised headache trigger. The NHS names it among its ten headache triggers and puts the mechanism plainly: “Poor posture causes tension in your upper back, neck and shoulders, which can lead to a headache.” The NHS page on headaches also lists bad posture among common causes, alongside stress, dehydration, skipped meals and taking too many painkillers.
It gets more interesting the more specific the page gets. The NHS page dedicated to tension headaches gives the causes as stress, sleep problems and caffeine. Posture is absent. The word does not appear anywhere on that page, though it does suggest a low, firm pillow and heat or cold packs “if you have neck pain and headaches”.
So the same organisation names posture on its general symptom page and on a triggers list that also includes ice cream, then leaves it off the clinical page for the headache type most desk workers have. Read that as a confidence gradient rather than a contradiction. The same gradient runs through the research below.
Two headache types are in play here, and they behave differently.
Tension-type headache is the common one. Both sides, a pressing or tightening band, no nausea, no light sensitivity. It’s also the most prevalent headache disorder on the planet. A 2025 meta-analysis in The Journal of Headache and Pain, pooling individual data from 41,614 adults across 17 countries, put the one-year prevalence at 33.2%. A third of adults, in any given year.
Cervicogenic headache is the rarer, more specific one. Pain that genuinely originates in the neck and refers up into the head. StatPearls describes it as “unilateral pain that starts in the neck and is referred from bony structures or soft tissues of the neck,” and puts its prevalence among headache patients at 0.4% to 4%.
Desk posture is genuinely involved in both. It is also rarely the whole story, whatever the wellness internet says. Worth understanding why.
How a Neck Turns Into a Headache
The connection isn’t vague. There’s a specific piece of anatomy that makes neck problems feel like head problems.
Sensory nerves from the top three vertebrae in your neck don’t run off to their own separate corner of the brain. They converge with the trigeminal nerve, which handles sensation for your face and much of your head, at a shared relay station in the upper spinal cord. The 2012 review Understanding Cervicogenic Headache in Anesthesiology and Pain Medicine calls this “the convergence of sensory inputs at the trigeminocervical nucleus.”
Signals arriving from your neck and signals arriving from your forehead land on the same neurons. The brain, reading the output, cannot always tell which one it came from. So an irritated joint at C2 gets experienced as pain behind the eye, which is why a problem in your neck can produce a headache with no neck pain attached to it at all.
Sitting in the loop is a set of four small muscles at the base of the skull, the suboccipitals. StatPearls notes that they “serve as postural support of the head and neck” and that they “have been investigated as a potential cause of cervicogenic headaches.” Their job is holding your head up. When your head drifts forward of your shoulders, they hold it up against a longer lever, all day.
There’s a further wrinkle that’s still being worked out. Those muscles connect directly to the covering of the spinal cord through a structure called the myodural bridge. A 2021 paper in Scientific Reports describes how it “connects the suboccipital musculature to the spinal dura mater,” and the authors are careful about what it means: “Clinically, dysfunctional suboccipital musculature was observed to be related to various types of headaches. However, its mechanism is still unknown.”
So there is a plausible anatomical route running from a tired neck muscle to a pain-sensitive membrane, seen clinically and not yet explained.
Nothing here requires a slipped disc or an injury. Small muscles, held under low load, for hours, in a position they weren’t designed to hold. Our guide to forward head posture covers what that position does to the rest of the neck.
Tension-Type or Cervicogenic? Telling Them Apart
The two respond to different things, so it’s worth a minute working out which you have.
| Tension-type headache | Cervicogenic headache | |
|---|---|---|
| Sides | Both, usually | One side, and it stays on that side |
| Feel | Pressing, tightening, a band | Starts in the neck, spreads to the front |
| Where it starts | Diffuse, no obvious origin | Base of the skull, then travels |
| Neck movement | Little effect | Makes it noticeably worse |
| Neck range | Usually normal | Often reduced |
| Nausea, light sensitivity | Absent | Absent (this is how it differs from migraine) |
| How common | 33.2% of adults per year | 0.4-4% of headache patients |
Harvard Health describes the cervicogenic pattern precisely. The pain “starts at the back of the neck at the base of the skull and radiates to the head and face”, usually reaching the forehead, sometimes behind the eyes, sometimes as far as the shoulder and arm.
Cleveland Clinic adds the test that separates it from migraine. A cervicogenic headache “may feel like a migraine with one-sided head pain. But you won’t have additional symptoms common with a migraine, like light sensitivity, sound sensitivity, nausea or vomiting.”
If turning your head or holding a phone to your ear reliably makes it worse, and it’s the same side every time, that leans cervicogenic. A symmetrical vice-grip that builds through the afternoon is more likely tension-type. Plenty of people get both, which is why this is a steer and not a diagnosis.
What the Research Actually Shows
Most articles on this topic overstate their hand at exactly this point, which is a shame, because the real picture is more useful than the confident one.
The association between a forward-carried head and headache is real and measurable. A 2025 cross-sectional study in the Journal of Oral & Facial Pain and Headache assessed 117 people with forward head posture and found the frequency of cervicogenic headache among them was 53.8%. More than half. The craniovertebral angle, which is the standard measurement of how far forward the head sits, was lower in the headache group: 38.95 ± 4.65 degrees against 41.11 ± 5.97 in the group without (p = 0.030).
Two things independently predicted headache in that sample. A decreased craniovertebral angle (OR 0.878, p = 0.014), and poor sleep quality on the PSQI (OR 1.140, p = 0.025). The authors flag their own limits: “other unmeasured variables may also contribute to CGH risk, indicating a multifactorial etiology.”
That second predictor deserves more attention than it usually gets. Sleep quality showed up alongside neck posture in the same model. If you’re chasing a chair while sleeping badly, you may be working on the wrong variable.
Now the part that gets left out.
The broader evidence linking forward head posture to neck symptoms is inconsistent, and it depends heavily on who you measure. A 2019 systematic review and meta-analysis in Current Reviews in Musculoskeletal Medicine pooled 15 cross-sectional studies and opened by acknowledging that “despite claims that FHP may be related to neck pain, this relation seems to be controversial.” In adults, ten studies produced a mean difference of 4.84 (95% CI 0.14 to 9.54), a significant gap, and correlations with pain intensity (r = -0.55) and disability (r = -0.42). In adolescents, the same comparison gave -1.05 (95% CI -4.23 to 2.12). Nothing. The authors concluded that age “played an important role as a confounding factor.”
A 2018 case-control study in the Brazilian Journal of Physical Therapy went further, comparing 32 people with chronic neck pain against 35 without, and reported that “FHP is neither different between CNP and asymptomatic participants nor correlated with either muscles size or endurance and clinical characteristics of these patients.”
So the posture-causes-pain story is not settled, and cross-sectional studies can’t sort out which came first anyway. A head that sits forward might drive the headache, or a sore neck might make someone carry their head differently in the first place. One photograph at one moment in time cannot separate those.
Which leaves an obvious question. If the observational evidence is this murky, what should you actually do about a headache?
What the Trials Show
Look at the trials instead of the snapshots and the picture sharpens considerably. When researchers get office workers moving their necks and shoulders, headaches go down. Repeatedly, in decent-sized randomised trials.
The NEXpro trial (2025). Published in The Journal of Headache and Pain, this stepped-wedge cluster RCT put 120 office workers through 12 weeks of neck and shoulder-girdle exercise, three times a week, 20 minutes a session, at their workplace, plus health-promotion sessions on sleep and stress. Headache occurrence dropped with an odds ratio of 0.46 (95% CI 0.25 to 0.84). Headache frequency in the first intervention period fell to an incidence rate ratio of 0.57 (95% CI 0.44 to 0.74). At baseline, 88 of the 120 had experienced at least one headache in the preceding four weeks, averaging 3.58 headache days per four-week block.
The strength-training trial (2014). In BioMed Research International, 351 office workers did 20 weeks of dumbbell work for the neck and shoulders, three 20-minute sessions a week. One hour a week, total. Headache intensity fell by 1.1 ± 0.2 points against the reference group (P < 0.001). What makes this one useful is the comparison built into it. One group was fully supervised, the other got minimal initial instruction at the start and then got on with it, and both landed on exactly the same number. Supervision changed nothing about the result.
The workplace education trial (2012). The largest of the three, in PLoS One, covered 1,881 workers with brief neck and shoulder exercises, a relaxation exercise, and instruction on not clenching through the day. The headache responder rate, meaning a 50% or better reduction, came out at a risk ratio of 1.58 (95% CI 1.28 to 1.92). Mean headache frequency fell by 1.72 days a month (95% CI -2.40 to -1.04).
Across all three trials, nobody measured whether anyone’s craniovertebral angle changed. The intervention was movement and load. People got better without anyone checking whether their posture had been corrected, which is an odd gap once you notice it.
That reframes the practical advice considerably. Chasing a perfect neutral spine is chasing a variable the successful trials didn’t even track. Interrupting long static holds, and giving the muscles that hold your head up some actual capacity, is what those trials did.
Harvard Health reports a similar figure for the cervicogenic end: “72% of people who completed six weeks of physical therapy reduced their headache frequency by 50% or more after a year.”
Is It Actually Your Posture?
Before you rearrange your desk, rule out the things that produce an identical afternoon headache. Each of these appears on the NHS list alongside posture, and each is easier to fix.
Painkillers. The NHS names “taking too many painkillers” as a cause of headaches in its own right. Regularly medicating a daily headache can sustain the headache. If you’ve been taking something most days for months, this needs raising with a GP before anything else on this list.
Sleep. The forward-head study above found poor sleep quality independently predicted cervicogenic headache in the same model as neck posture. If you’re sleeping badly, that’s a live variable in your own headaches too.
Dehydration and skipped meals. Both on the NHS list. A headache that arrives after you worked through lunch has an obvious first suspect.
Caffeine. The NHS names caffeine on its tension headaches page, and a three-o’clock headache in someone whose last coffee was at eight in the morning fits withdrawal at least as well as it fits posture.
Eye strain. Eyesight problems appear on the NHS causes list. If you haven’t had your eyes tested in a few years and you spend nine hours on screens, get that done before you buy a monitor arm.
Stress. Top of the NHS list for tension headaches. Worth noting that the successful trials above included relaxation and stress-management components, not just exercise.
A posture-shaped headache builds through a working day rather than being there when you wake up. It’s worse on heavy desk days and better at weekends, it arrives with neck or upper-back tightness, and it tends to ease within an hour of getting away from the desk. If yours is worst first thing in the morning, or shows no relationship at all to how much you sat, look elsewhere first.
When to Stop Self-Diagnosing
Most desk-work headaches are dull and annoying and nothing worse. A small number are not, and the difference is worth knowing cold rather than looking up at eleven at night.
The NHS says to call 999 or go to A&E if a headache comes with a seizure, numbness or weakness, sudden extreme pain, a recent head injury, difficulty speaking, walking, balancing or remembering, drowsiness or confusion, loss of vision, a rash that doesn’t fade under a glass, or a very high temperature with a stiff neck or dislike of bright light.
Short of that: see a GP if you’ve tried treating it yourself and it isn’t getting better or is getting worse, or if you regularly get headaches. A daily headache for six months is a GP conversation, not an ergonomics project.
What Actually Helps at a Desk
This list is weighted towards what the trials above actually did.
Break the static hold before it accumulates. The muscles holding your head up fatigue with duration, not with badness of posture. The trial evidence is about interrupting long periods, and the practical version is picking an interval you’ll honour. The 20-8-2 rule is a ready-made structure if you’d rather not invent one.
Give the neck and shoulders some load. This is the intervention with the most direct evidence. One hour a week of dumbbell work, split into three sessions, dropped headache intensity by a full point in 351 office workers, and supervision turned out not to matter. Shrugs, lateral raises, front raises, reverse flies. Our posture exercises for desk workers covers the desk-friendly end of it, and the text neck exercises piece has the neck-specific work.
Raise the screen. A laptop flat on a desk makes forward head posture structurally unavoidable, no matter how much you intend to sit up. The American Migraine Foundation recommends positioning your screen at or slightly below eye level and bringing handheld devices up rather than dropping your head to them. Our guide to monitor height and neck pain gets into the specifics.
Deal with the sleep. Poor sleep quality was an independent predictor in the forward-head study, and it’s a variable most people ignore while buying lumbar cushions.
Move the suboccipitals directly. Cervicogenic pain answers to work aimed at the top of the neck, which general fitness tends to miss. Harvard notes that craniocervical flexion, gently nodding the chin against light resistance, builds stability in exactly the region involved. Chin tucks are the desk version, and they take about thirty seconds.
Get assessed if it’s one-sided. Physical therapy has real numbers behind it for cervicogenic headache, and if your pain is consistently on one side and provoked by neck movement, that’s the referral worth asking for.
What Helps Less Than You Think
Sitting perfectly upright. Holding an idealised position is still holding a position, and the trials that reduced headaches never verified anyone’s alignment. Rigid uprightness fatigues fast and tends to collapse into a worse slump than you started from.
A better chair on its own. A good chair makes a long hold more tolerable without making it any shorter, and length is the thing that matters. Chairs get bought as a way of skipping the movement conversation, and movement is what shifted the numbers in every trial above.
Stretching without loading. Stretching a tight upper trapezius feels good for about twenty minutes. The trials that produced lasting change used strength and endurance work. Stretch if you like it, but don’t expect it to be the intervention.
Taking a painkiller every day and carrying on. The NHS lists painkiller overuse among the causes of headache, which makes daily medication a plausible reason a daily headache never lifts. If you’ve been on something most days for months, that’s a GP conversation.
Nobody sits down intending to hold their head forward for two hours. It happens because attention goes to the work and not to the body, which makes noticing it the awkward bit. That specific gap is what SitApp exists to fill: the Droid watches your posture through your webcam, entirely on your own machine, and nudges you when you’ve been slumped or static too long. No images ever leave your computer. If your headache is time-and-position dependent, having something external tracking the time beats intending to remember.
Frequently Asked Questions
Can bad posture cause headaches every day? It can contribute to daily headaches, though a genuinely daily headache is worth a GP appointment rather than a desk overhaul. Chronic daily headache has several causes, and painkiller overuse is a common and treatable one. The NHS advises seeing a GP if you regularly get headaches or if self-treatment isn’t working.
Where does a posture headache hurt? Typically at the base of the skull and up over the back of the head, often spreading to the forehead or behind the eyes. Tension-type headaches are felt on both sides as a pressing or tightening band. Cervicogenic headaches start in the neck and stay on one side, and they get worse with neck movement.
How long does it take for a posture headache to go away? An individual episode usually eases within a few hours of getting away from the desk. Changing the pattern takes longer. The office-worker trials ran 12 to 20 weeks, and the biggest headache reductions showed up over that timescale rather than in the first week.
Can fixing my posture cure my headaches? It can reduce them meaningfully, and the trial evidence supports that. Whether “fixing posture” is the right description is less clear, because the successful trials used exercise and movement breaks and never measured whether anyone’s alignment actually changed. Adding capacity and interrupting long static holds is what produced the results.
Is a headache from bad posture the same as a migraine? No. Migraine typically brings nausea, light sensitivity or sound sensitivity, and posture-related headaches don’t. Cleveland Clinic uses exactly that distinction to separate cervicogenic headache from migraine. That said, people with migraine do show more neck dysfunction than the general population, so the two often travel together.
Why do I only get headaches on work days? That pattern points fairly strongly at something in the working day: sustained sitting, screen height, stress load, caffeine timing, or skipped meals. A headache that’s reliably absent at weekends and present by Tuesday afternoon is worth mapping against those before assuming it’s posture.
The Short Version
Bad posture can cause headaches. The NHS lists it as a trigger, the anatomy connecting neck to head is well described, and more than half of people with forward head posture in one 2025 study had cervicogenic headache.
The evidence that your specific alignment is the culprit is weaker than the internet lets on. Forward head posture separates adults with and without neck pain but not adolescents, and one case-control study found no difference at all. Cross-sectional studies cannot tell you which came first.
The evidence for what fixes it is much stronger. Three randomised trials covering over 2,300 office workers all reduced headaches with neck and shoulder exercise plus movement breaks, and none of them checked anyone’s posture afterwards.
This week, three things. Raise your screen so the top sits at or just below eye level. Pick a break interval and let something other than your memory enforce it. Start three 20-minute sessions a week of shoulder and neck work, because that’s the dose that dropped headache intensity by a full point in 351 people.
Then rule out the cheap stuff before you spend money on the expensive stuff: sleep, water, lunch, caffeine timing, an eye test, and how many painkillers you’ve been taking. And if the headache is one-sided, provoked by turning your head, or simply not shifting, book the appointment.
This article is general information, not medical advice. Headaches have many causes and some of them need proper assessment. If your headaches are severe, persistent, changing in pattern, or getting worse, see a doctor.
