What the dew point actually is
Air holds water vapour, and there is a ceiling on how much it can hold at a given temperature. Warm air has a high ceiling and cold air a low one — that relationship is steep, roughly a doubling of capacity for every 10 °C. Relative humidity is not an amount of water; it is the fraction of that ceiling currently being used. Fifty percent at 30 °C is considerably more water than fifty percent at 10 °C.
The dew point is the temperature to which that particular parcel of air would have to be cooled, without adding or removing any water, for the ceiling to drop to what the air already holds. At that point relative humidity reaches 100% and water starts coming out. It follows that the dew point is a direct measure of absolute water content: two rooms with the same dew point contain the same amount of vapour per cubic metre, whatever their thermometers say and whatever their humidity readings say.
That property is what makes it the useful number for building work. Relative humidity changes the moment air moves from a warm room to a cold one; the dew point does not. A surface anywhere in the building that is below the dew point of the air touching it will collect water, and no amount of ventilation with air of the same dew point will change that.
The formula, stated plainly
Dew point cannot be written in closed form from first principles, so every calculator uses an approximation to the saturation vapour pressure curve. This one uses the Magnus form with the coefficients from Alduchov and Eskridge (1996):
γ = (17.625 × T) / (243.04 + T) + ln(RH / 100)
Td = (243.04 × γ) / (17.625 − γ)
with T and Td in degrees Celsius. Those coefficients were fitted to minimise error against the reference formulation, and the paper reports a maximum error under about 0.1 °C across roughly −40 to +50 °C, which comfortably covers everything a building or a shop will present. Below freezing the physical situation changes — saturation over ice differs from saturation over supercooled water — and the approximation drifts. It is an approximation. It is a very good one in the range that matters, and it is not an exact answer.
The absolute humidity figure alongside it comes from the same saturation curve: vapour pressure converted to grams of water per cubic metre with the ideal gas relation. At 20 °C and 100% relative humidity that gives 17.3 g/m³, which matches published tables.
Where the number is used in building work
The first use is diagnosing condensation that has already appeared. Take the air temperature and humidity in the room, get the dew point, and take a surface temperature with an infrared thermometer. If the surface reads below the dew point, the mechanism is settled and the argument about whether it is a leak or condensation is over. Cold spots at the junction of a wall and a floor, behind furniture pushed against an outside wall, and at window reveals are the usual sites, because those are where the surface temperature falls furthest below the room.
The second use is coating work. Applying paint, sealer or coating to a surface at or near the dew point puts a film over condensed moisture, and finishes fail that way in ways that only show up weeks later. Most product instructions state a minimum margin above the dew point — commonly around 5 °F or 3 °C — and the instruction on the tin is what governs, not this page. The point of the number here is that you can check the margin before you open the tin rather than after.
The third use is judging what a dehumidifier will achieve. A dehumidifier lowers the dew point of the air, which raises the margin at every surface simultaneously. Heating the room raises the surface temperatures but leaves the dew point where it was, which also works and costs more to run. Which of those makes sense depends on the building, and the dehumidifier sizing calculator and the humidifier sizing calculator take the equipment side of it.
Comfort, and why the dew point predicts it better than humidity
Human comfort tracks the dew point much more closely than it tracks relative humidity, because sweating works by evaporation and evaporation is governed by the vapour pressure difference between skin and air. As a rough guide that people find reliable in practice: a dew point below about 55 °F feels dry and pleasant, the low 60s feels noticeably humid, and above about 70 °F feels oppressive regardless of what the thermometer says. That is why a 90 °F day in a dry climate is tolerable and an 85 °F day on a humid coast is not, even though the second one may show a lower relative humidity reading.
The same logic explains winter indoor dryness. Cold outdoor air brought inside and heated keeps its low dew point while its relative humidity collapses, which is why heated rooms in winter feel dry and why static and cracked woodwork follow. Adding heat adds no water. For material movement that follows from all this, the wood movement calculator works in equilibrium moisture content rather than dew point, which is the right variable for timber.
Questions people ask
Can the dew point be higher than the air temperature?
Not in any stable situation, no. At 100% relative humidity the dew point equals the air temperature, and that is the ceiling. A reading that puts the dew point above the air temperature means one of the two measurements is wrong, most often a humidity sensor that has been wetted or has drifted out of calibration. Fog and mist are the edge case, where the air is at saturation and holding suspended droplets, and even there the dew point equals rather than exceeds the temperature.
Why does my cold drink sweat but my cold window does not?
Because the glass of a drink is far colder than the window and one of them is below the dew point. It is not about the material or the shape. Take the surface temperature of each with an infrared thermometer and compare both against the dew point figure here; the one that sweats will read below it. The same reasoning explains why the bottom corner of a window fogs while the middle stays clear, since the frame conducts and the corner runs colder.
How accurate is this?
The formula is the limiting factor least of all. The Magnus approximation with these coefficients is quoted as accurate to about a tenth of a degree C over the range that covers ordinary building and shop conditions. Your inputs are the weak link: relative humidity from an inexpensive sensor is often a few percent out, and a few percent moves the dew point by around half a degree C at room conditions. Temperature is usually the more reliable of the two readings.
Does the dew point change with air pressure?
Slightly, and for ordinary work you can ignore it. The saturation vapour pressure curve is a function of temperature, and the calculation here does not take pressure as an input, which is standard practice. At altitude or in a pressurised system the exact relationship shifts a little. If you are doing compressed air work, that is a case where pressure genuinely matters, because compressing air raises its dew point and that is why compressed air lines drop water.
What margin above the dew point should I keep before painting?
Whatever the product instructions say, and they do say. A figure around 5 degrees F above the dew point is commonly specified, but it varies by product and by the coating chemistry, and the instruction on the container is the one that governs the warranty. Use this page to check the margin you have, not to choose the margin you need. Remember that the surface temperature is what matters, not the air temperature, and that the two diverge most in the situations where you are most tempted to proceed anyway.