Does Dehydration Actually Cause Headaches?

Drink some water is the most common headache advice there is. It is cheap, it is harmless, and the evidence behind it is far thinner than the confidence with which people give it.

Someone says they have a headache. Someone else says: you're probably dehydrated, drink some water.

It's a reflex, and it's a reasonable one. Water is free, it carries no risk, and sometimes it works.

But go looking for the evidence and you find something more interesting than a clean yes. The only randomised trial ever run on this found that the measure people care about did not move. And in emergency departments, giving migraine patients fluid directly into a vein does not improve their pain.

Before you read on: this guide covers situations that can become medical emergencies. Read the scope and safety note first.

The Short Answer

Bottom line: dehydration can contribute to headache and rehydrating is cheap and low-risk, so it is worth trying. But the evidence that drinking water treats an existing headache is weak, and prevention is better supported than treatment.

Three findings that don't appear in the usual advice:

• The one trial split. Quality-of-life scores improved. Actual headache days did not.
• Intravenous fluids don't improve migraine pain in emergency settings, which is a hard test of the hydration theory.
• Nobody knows the mechanism. Harvard Health says so directly, despite the brain-shrinking explanation appearing on almost every page about this.

Is a Dehydration Headache a Real Thing?

Yes, but it sits in a vaguer position than most coverage suggests, and the classification detail is genuinely revealing.

The international reference for classifying headaches is the ICHD-3, which sorts them into named categories. Chapter 10 covers headache attributed to a disorder of homoeostasis, which is where a dehydration headache would logically live. Its named subcategories are: hypoxia and hypercapnia (including high-altitude, aeroplane travel, diving and sleep apnoea headaches), dialysis, arterial hypertension, hypothyroidism, fasting, cardiac cephalalgia, and a catch-all for other disorders.

Dehydration is not named anywhere in that chapter. Skipping meals has its own diagnostic category. Not drinking enough does not.

That is not a claim that dehydration headaches don't exist. A clinician would place one under the catch-all. But for a condition supposedly responsible for a large share of everyday headaches, the absence is telling.

The mechanism nobody can confirm

Nearly every article on this subject explains it the same way: dehydration makes brain tissue shrink slightly, pulling on the pain-sensitive membranes around it.

Harvard Health is refreshingly direct about this: "The exact reason why dehydration causes headaches is unknown." It offers the membrane-traction idea as one possibility and an exaggerated perception of pain as another, and it hedges both (possibly triggered, possible explanation, might feel).

So the confident mechanical story you have read a dozen times is a hypothesis presented as physiology. Worth knowing before you build much on it.

Does Drinking Water Actually Help?

One randomised trial has tested it. The result splits down the middle, and coverage tends to report only the flattering half.

Spigt and colleagues, in Family Practice (2012), randomised 102 people with recurrent headaches. Fifty continued as usual; 52 were instructed to drink an extra 1.5 litres of water per day.

Read the third row. The trial's own report is that drinking more water "did not result in relevant changes in days with at least moderate headache."

That is the outcome a person with headaches actually cares about: how many days does this ruin. It did not move. What moved was a self-reported questionnaire score and a self-reported impression of improvement.

Migraine-Specific Quality of Life is a validated instrument, not a throwaway. But it asks people how much headaches are affecting their life, which is a judgement. Headache days are a count.

Why That Result Needs Its Caveats

The trial was formally appraised, and the appraisal found problems that bear directly on which half of the result you trust.

Price and Burls, in the Journal of Evaluation in Clinical Practice (2015), examined the study and reported it as "underpowered" with "methodological shortcomings." Their most consequential finding:

"participants were partially un-blinded during the trial increasing the risk for bias"

Sit with what that means in combination. The outcome that improved was self-reported, in a trial where people knew which group they were in, after being told that drinking more water might help their headaches.

That is the weakest possible configuration for a subjective measure. It is exactly the arrangement in which expectation produces an effect, and it is why the objective measure staying flat matters so much.

To be fair to the appraisers, their overall verdict is balanced rather than dismissive. They describe water intake as "a cost effective, non-invasive and low-risk intervention", and conclude that "a larger methodologically sound randomized controlled trial is needed to confirm efficacy." Not disproven. Unproven, and worth trying anyway because trying costs nothing.

The harder test

There is a more demanding version of this experiment, and it has been run.

A 2021 review of dehydration and headache reports that intravenous fluids for migraine in emergency settings "have not been shown to improve pain outcomes."

Consider what that removes from the equation. No absorption question. No compliance question. Fluid delivered directly into the bloodstream, in a clinical setting, to people in significant pain. If a fluid deficit were driving the pain, this is where you would see it resolve.

The same review is careful not to overcorrect, and neither should we. It notes that maintaining good hydration does help treat orthostatic and "coat-hanger" headaches arising from autonomic disorders, and that preventing fluid losses can mitigate secondary headaches provoked by dehydration. Hydration is genuinely therapeutic in specific, identified situations. It is just not the general-purpose headache treatment it gets used as.

Dehydration and Migraine

Here the evidence runs in dehydration's favour, via a test that demolishes several more famous triggers.

Migraine has a premonitory phase: hours before the pain begins, brain changes are already underway and produce symptoms. A systematic review found at least one premonitory symptom in a pooled 29% of population-based studies and 66% of clinic-based studies (both with very wide confidence intervals and high heterogeneity, so treat them as broad indications). The most common were fatigue (49%), neck stiffness (46%) and mood change (37%).

This creates a trap for anyone identifying their own triggers. If the attack has already begun and it makes you crave chocolate, and you eat chocolate, and then your head hurts, you will conclude chocolate caused it. Researchers put the problem precisely:

"Early brain changes mediating premonitory symptoms before headache may, therefore, lead to the correct association of these symptoms with corresponding trigger factors, but incorrect attribution that the trigger factor is directly causing the headache."

They then tested which perceived triggers matched a corresponding premonitory symptom. Good agreement turned up for light with photophobia, sound with phonophobia, eating with food cravings, skipping meals with food cravings, and stress with mood changes. Several classic triggers look, on this evidence, like symptoms wearing a disguise.

Dehydration and thirst showed poor agreement. And thirst does not appear at all among the individual premonitory symptoms in the systematic review above.

So the confound that undermines chocolate and bright light does not appear to undermine dehydration. That is a point in favour of it being a real trigger rather than a misread symptom, and it is the opposite of the tidy contrarian conclusion this section could have reached.

What To Actually Do

Treat it as cheap prevention with a low ceiling, rather than as a treatment.

• Drink water when you have a headache. It costs nothing, carries no risk, and might help. Just calibrate what you expect from it.
• Give it an hour or two. Harvard Health suggests a water-deprivation headache should resolve within an hour or two of drinking 16 to 32 ounces. If yours doesn't, fluid probably wasn't the cause, and treating it as one delays finding what was.
• Weight prevention over rescue. The evidence for preventing fluid losses is better than the evidence for reversing a headache in progress. Steady daily intake beats emergency rehydration.
• Check whether you're actually dehydrated rather than assuming. Our guide to dehydration signs that don't involve thirst covers what the real signals are, and urine colour has its own limits worth knowing.
• Work out your baseline if headaches are frequent. The hydration calculator sets a daily figure from your own body and activity.
• Track it honestly. Log headache days, not how you felt about them. That's the distinction the trial turned on, and it applies just as well to your own experiment on yourself.

No product recommendation on this page. Search this question and you'll find electrolyte brands concluding, with some enthusiasm, that electrolytes help headaches. The evidence above doesn't support a product claim: the one trial tested plain water, and the harder test used intravenous fluid. Selling you something here would put this page in the group it's describing.

When a Headache Isn't About Hydration

Some headaches need attention rather than a glass of water. These are the ones not to reason your way past.

Seek urgent medical care for:

• A headache that comes on suddenly and severely, reaching maximum intensity within seconds or minutes
• Headache with fever, neck stiffness, or a rash
• Headache with confusion, weakness, numbness, vision loss, or difficulty speaking
• Headache after a head injury
• Headache with a seizure, or with loss of consciousness

And see a clinician, without urgency but without indefinite delay, about a headache that is new for you, that is steadily worsening over days or weeks, that wakes you from sleep, that is clearly different from your usual pattern, or that has you reaching for painkillers on most days.

That last one matters and rarely gets mentioned alongside hydration advice: frequent painkiller use can itself perpetuate headaches. If you're medicating often, that's a conversation worth having rather than a hydration problem to solve.

FAQ

Can dehydration cause a headache? It can, though it more often worsens something already there. A 2021 review in Current Pain and Headache Reports puts it as dehydration alone may cause headache, but oftentimes exacerbates underlying medical conditions. Notably, dehydration is not a named diagnostic category in the international headache classification, while fasting is.

Does drinking water actually get rid of a headache? Less reliably than the advice implies. The only randomised trial gave 102 patients an extra 1.5 litres a day. Quality-of-life scores improved significantly, but the trial found no relevant change in the number of days with at least moderate headache. The measure that improved was self-reported, in a trial where participants were partially unblinded.

How long does a dehydration headache take to go away? Harvard Health suggests a water-deprivation headache should resolve within an hour or two after drinking 16 to 32 ounces of water. If a headache does not respond to rehydration in that window, it is worth considering that fluid was not the cause.

Is dehydration a real migraine trigger? It holds up better than several more famous triggers. Research testing whether perceived triggers are actually early symptoms of the attack found good agreement for light, sound, food and stress, but poor agreement for dehydration and thirst. That points toward dehydration being a genuine trigger rather than a misread symptom.

When is a headache something more serious? Seek urgent care for a headache that comes on suddenly and severely, one accompanied by fever, neck stiffness, rash, confusion, weakness, vision loss or difficulty speaking, one following a head injury, or any headache that is clearly different from your usual pattern. Those are not hydration problems.

Keep Reading

• The Signs of Dehydration That Don't Involve Feeling Thirsty helps you check whether you're actually dehydrated rather than assuming it.
• Urine Colour and DIY Hydration Checks: How Accurate Are They? covers the self-check most people reach for, and how far it can be trusted.
• Do Electrolyte Drinks Actually Help With Hangovers? covers hangover headache, which runs on a different mechanism from this one.

A Note on Scope

This article is for informational purposes only and isn't medical advice. Headaches have many causes, some of them serious. Seek urgent medical care for a headache that begins suddenly and severely, or that comes with fever, neck stiffness, rash, confusion, weakness, vision changes or difficulty speaking, or that follows a head injury. See a clinician about any headache that is new, changing, worsening over weeks, or different from your usual pattern, and about frequent headaches or escalating painkiller use. Nothing here is a substitute for a diagnosis.

Sources

• Spigt M, et al. "A randomized trial on the effects of regular water intake in patients with recurrent headaches." Family Practice, 2012;29(4):370 supplies the arm sizes, the 1.5 litre intervention, the quality-of-life improvement with confidence interval, the 47% versus 25% perceived-improvement figures, and the null result on headache days.
• Price A, Burls A. "Increased water intake to reduce headache: learning from a critical appraisal." Journal of Evaluation in Clinical Practice, 2015;21(6):1212-1218 supplies the underpowered and methodological-shortcomings findings, the partial unblinding, the low-cost and low-risk verdict, and the call for a larger trial.
• Arca KN, Halker Singh RB. "Dehydration and Headache." Current Pain and Headache Reports, 2021;25(8):56 supplies the finding that intravenous fluids have not been shown to improve migraine pain outcomes, the may-cause-but-often-exacerbates framing, the orthostatic and coat-hanger headache exception, and the prevention point.
• "Are some patient-perceived migraine triggers simply early manifestations of the attack?" supplies the premonitory-misattribution argument and quotation, the trigger and symptom pairs showing agreement, and the poor agreement found for dehydration and thirst.
• "Premonitory symptoms in migraine: a systematic review and meta-analysis" supplies the pooled premonitory prevalences and the individual symptom frequencies.
• Harvard Health, "Can dehydration cause headaches?" supplies the statement that the mechanism is unknown, the two hedged candidate explanations, and the resolution time after drinking 16 to 32 ounces.
• ICHD-3, Chapter 10: Headache attributed to disorder of homoeostasis supplies the full list of named subcategories, in which dehydration does not appear and fasting does.

Not used: the brain-shrinkage mechanism as an established explanation. It appears on nearly every page covering this topic, but no primary source establishing it was located, and Harvard Health describes the mechanism as unknown. It is presented here as a hypothesis rather than repeated as fact.