Article

What barometric pressure causes headaches?

No particular barometric pressure reading — 30.00 inHg (1015.9 hPa) included — has been linked to headaches in the research. What studies associate with migraine, in some people, is pressure falling, and the size that recurs is about 0.18–0.30 inHg (6–10 hPa); effect sizes are modest, findings are mixed, and sensitivity varies widely between people.

Level or change: does 30 inHg matter?

It does not. A reading of 30.00 inHg (1015.9 hPa) sits comfortably inside the normal band on this site’s pressure zones, which runs from 29.68 to 30.12 inHg (1005–1020 hPa). It is an ordinary, roughly average sea-level value — Denver’s ten-year mean is 29.97 inHg (1014.8 hPa) — and none of the studies below links a static level, high or low, to headache onset.

The question people actually type assumes there is a number on the dial at which headaches start. The research does not look for one. It looks at how much the pressure moved and in which direction. A still day at 30.20 inHg (1022.7 hPa) and a still day at 29.70 inHg (1005.8 hPa) are, as far as this literature can tell, equally unremarkable. A day on which pressure slides from one to the other is the kind of day the studies are about.

If you arrived with a reading in hand and want to know where it sits, normal barometric pressure covers the range, and is 29.95 high or low has a lookup table. If the units themselves are the confusing part — studies quote hPa, most US readings are in inHg — what is barometric pressure explains them. Here, every value is given in both: 1 hPa = 0.02953 inHg.

What the studies found

The evidence comes almost entirely from headache diaries: people with migraine record their attacks, and researchers line the dates up against weather records. That design can show association. It cannot show that pressure produced any particular headache.

The studies that found a link

The number most often quoted comes from a Japanese diary study by Okuma, Okuma and Kitagawa, published in SpringerPlus in 2015. Patients recorded the atmospheric pressure when a migraine began. Taking 29.91 inHg (1013 hPa) as standard atmospheric pressure, the authors found attacks clustered as a low approached and pressure had fallen to 29.62–29.74 inHg (1003–1007 hPa) — that is, 0.18–0.30 inHg (6–10 hPa) below standard. Two details matter. The figure is a fall measured against a fixed reference, not a 24-hour change. And 1003–1007 hPa straddles the bottom edge of the normal band; it is not an exotic low.

An earlier study from Japan, Kimoto et al. in Internal Medicine (2011), followed 28 people with migraine. Weather change was associated with migraine development in 18 of them, and 14 identified low pressure specifically. The authors read this as pressure change being one of several factors linked to worse migraine in their patients — a cautious conclusion from a small group.

The studies that found little or nothing

The larger, more statistically careful studies are less encouraging. Zebenholzer et al., writing in Cephalalgia in 2011, followed 238 migraine patients in Vienna through 90-day diaries against 11 weather measurements and 17 weather types. A few signals appeared — none of them a pressure fall — and none survived correction for multiple testing. Their conclusion: “The influence of weather factors on migraine and headache is small and questionable.”

Hoffmann et al. tracked 100 migraineurs in Berlin for a year, in Annals of Clinical and Translational Neurology (2015). Pooled together, the group showed no significant association with pressure, temperature or humidity. Tested one by one, 13 of the 100 were clearly weather-sensitive. The authors concluded that predicting an individual attack from weather data was not possible. The same group’s smaller 2011 study of 20 patients described the evidence in its abstract as “inconclusive” and found a strong association in 6 — driven by lower temperature and higher humidity rather than pressure.

Belief versus measurement

Prince et al., in Headache (2004), asked 77 migraineurs whether weather affected them before checking their calendars against the weather: 62.3% said yes. The data found 26 people (33.7%) sensitive to the first of three combinations of weather variables, with smaller groups of 11 (14.3%) and 10 (12.9%) sensitive to the other two. The study supports an influence of weather in some people, but more of them believed in it than could be demonstrated.

Put together, the share of migraine patients who look weather-sensitive runs from 13% under the strictest test to a third or more in the most generous, and the largest study found no reliable weather effect at all. There is no single settled percentage.

High or low pressure?

Where studies find anything about pressure, it is mostly on the falling side. Okuma’s attacks clustered as pressure dropped toward an approaching low, and Kimoto’s patients most often named low pressure. That is why this site counts drop days separately: falls are the direction most often described in pressure-sensitivity reports.

It is not unanimous. Zebenholzer’s Vienna study found a hint in the other direction — a ridge of high pressure went with more headache days (headache in general, not migraine) — though, like everything else in that study, it did not hold up after correction. The honest summary: most studies that find a pressure signal find it in falls, one careful study hints at the opposite, and others find nothing reliable at all.

As for why a fall would matter, no mechanism is established. Explanations about blood vessels or pressure in the sinuses and inner ear appear mainly on commercial sites and, at most, as proposals in the research — including Okuma’s own discussion. Pressure changes also arrive together with shifts in temperature, humidity and wind, which diary studies struggle to separate.

How often a 6 or 10 hPa change happens

This is where the site’s own data adds something. The outlook calendar on every city page bins each day by its largest 24-hour pressure change at exactly the sizes the literature uses: under 0.18 inHg (6 hPa), 0.18–0.30 inHg (6–10 hPa), and 0.30 inHg (10 hPa) or more. The methodology page defines these as swing days and big-swing days and is clear that they are conventional bins, not clinically validated thresholds. Ten years of Open-Meteo ERA5 hourly reanalysis, 2015–2024, show how often each one comes around:

The spread is the point. A change at the top of the site’s conventional bins — 0.30 inHg (10 hPa) in 24 hours — arrives on about 43% of days in Boston and about 3 days a year in Miami. In the typical city of the 50 it comes 86.1 times a year, roughly one day in four. A threshold that is crossed on four days in ten in Boston cannot, on its own, account for any one person’s headaches there, which is part of why population-level figures say so little about an individual.

Two caveats. First, the 6 hPa count saturates in volatile interior cities: Denver’s 281.6 swing days are about 77% of the year, so the 10 hPa big-swing count is the figure that actually tells cities apart. Second, these are sea-level-reduced values, and that reduction inflates apparent variability at altitude. Denver sits at 1,615 m and Colorado Springs at 1,832 m, so part of the Front Range’s place at the top of this table reflects elevation, not only weather.

What to do with this

The one thing this site suggests is a comparison, not a treatment. Keep a simple symptom log, then line it up against the 30 days of pressure history on your city’s outlook calendar — on any of the 50 city pages, or for any US location if your town is not one of them. Find the bad days and see what the barometer was doing. Over a month or two that tells you far more about whether you track pressure than any study average, and it can just as easily show that you do not.

Pressure change has been statistically associated with migraine onset in some people, and not in others; effect sizes are modest, findings are mixed, and individual sensitivity varies widely. Nothing on this page predicts a headache, diagnoses anything, or is medical advice. If migraine is affecting your life, that is a conversation for a clinician.

All 50 city pages · Check any US location · Methodology