Temperature is the one conversion where you cannot just multiply, because the scales disagree about where zero is. Each conversion therefore has both a scale factor and an offset:
F = C × 9/5 + 32C = (F − 32) × 5/9K = C + 273.15 — same size of degree, different zeroF = (K − 273.15) × 9/5 + 32The 9/5 exists because there are 180 Fahrenheit degrees between the freezing and boiling points of water where Celsius has 100. The 32 exists because Fahrenheit put zero at the freezing point of a brine solution rather than of water. Kelvin drops the offset problem by starting at absolute zero, the temperature at which molecular motion reaches its minimum, so there are no negative kelvin values.
Convert 37 C, normal human body temperature: 37 × 9/5 = 66.6, plus 32 = 98.6 F. That famous figure is a conversion artefact — the original measurement was 37 C, a rounded average, so 98.6 carries a precision that never existed. Modern studies put the average closer to 36.6 C, or 97.9 F.
Going back: 350 F, a common oven setting. 350 − 32 = 318, × 5/9 = 176.67 C. Every cookbook writes this as 175 or 180, because rounding to the nearest 5 is well inside what an oven thermostat can hold.
There is exactly one temperature that reads the same in Celsius and Fahrenheit, and it is −40. Check it: −40 × 9/5 = −72, plus 32 = −40. Below that point Fahrenheit numbers are higher than Celsius ones; above it they are lower. It is a useful sanity check — if you convert a very cold temperature and the Fahrenheit value came out smaller than the Celsius value, you have made a sign error.
| Celsius | Fahrenheit | Kelvin | Reference point |
|---|---|---|---|
| −273.15 | −459.67 | 0 | Absolute zero |
| −40 | −40 | 233.15 | The scales cross |
| 0 | 32 | 273.15 | Water freezes |
| 21 | 69.8 | 294.15 | Comfortable room |
| 37 | 98.6 | 310.15 | Body temperature |
| 100 | 212 | 373.15 | Water boils at sea level |
This is the error that shows up in engineering and weather reporting. If a forecast says the temperature will rise by 10 C, that is a difference, and differences ignore the offset entirely:
Change in F = change in C × 9/5
So a rise of 10 C is a rise of 18 F, not 50 F. Applying the +32 to a difference is meaningless — you would be adding the freezing point of water to a gap. The same rule applies to tolerances on a spec sheet, temperature coefficients, and any figure phrased as per degree. A material expanding 12 micrometres per metre per degree Celsius expands 6.67 per degree Fahrenheit, obtained by dividing by 1.8 with no offset anywhere.
An air temperature on its own does not tell you how it will feel. Humidity, wind and sunlight change the experience substantially, which is why weather services publish a heat index or wind chill alongside the raw number. Those are separate models with their own formulas — this converter deals only in the measured temperature, which is the honest limit of a pure unit conversion.
Just kelvin, written 300 K with a space and no degree symbol. It is an absolute scale, so the word degree is dropped by convention.
Inertia rather than merit. Fahrenheit does give finer whole-number resolution for weather — a degree F is smaller than a degree C — which is the one practical argument in its favour. For science and cooking chemistry, Celsius and kelvin are simpler because water and absolute zero sit at round numbers.
An absolute scale using Fahrenheit-sized degrees: R = F + 459.67. It survives in some American thermodynamics and aerospace work. You will rarely meet it outside those fields.
No. The conversion is arithmetic performed in your browser, with nothing transmitted or stored.
Every tool comes with a written guide, and every category is one click away.