Key takeaways
- Celsius and Fahrenheit have different zero points and different degree sizes. One Celsius degree equals 1.8 Fahrenheit degrees.
- Kelvin shares the Celsius step size but starts at absolute zero, so K = °C + 273.15.
- Temperature differences convert differently. A 10 °C change is an 18 °F change, not a 50 °F change.
- Use 273.15, not 273, when precision matters; the 0.15 K error is small but real.
- −40 is the crossover point where the two scales read the same number.
How the two everyday scales are built
Both scales were originally anchored to two fixed points of water at standard atmospheric pressure. Celsius put the ice point at 0 and the steam point at 100, for 100 intervals. Fahrenheit put them at 32 and 212, for 180 intervals. The ratio 180/100 = 9/5 sets the step size, and the 32-degree offset sets the zero. Solving those two facts gives °F = °C × 9/5 + 32 and its inverse °C = (°F − 32) × 5/9.
A 20 °C room is 68 °F; a 98.6 °F body temperature is 37.0 °C; a 72 °F day is 22.2 °C. The formula is short enough to do in your head in two steps: subtract 32, multiply by 5, divide by 9. The common failure is skipping the subtraction, which is the offset that makes the scales different in the first place.
The crossover is worth remembering. Set °F = °C and solve: C = 1.8C + 32, so −0.8C = 32 and C = −40. At −40 the two scales agree, which makes it a useful sanity check for a conversion routine.
Kelvin and the absolute scale
Kelvin begins at absolute zero, the lowest thermodynamic temperature, where a system has minimum possible energy. Absolute zero is 0 K, −273.15 °C, and −459.67 °F. Because the kelvin is the SI base unit of thermodynamic temperature, it is the scale used in physical science and in most engineering calculations involving gas laws or heat transfer.
Since 2019 the kelvin has been defined by fixing the Boltzmann constant at exactly 1.380649 × 10⁻²³ J/K, rather than by the triple point of water. The size of the kelvin is unchanged; only the definition moved. One kelvin is exactly the same size as one degree Celsius, so a temperature difference of 5 °C is also a difference of 5 K.
Style matters in technical writing: 300 K, not 300 °K and not 300 degrees Kelvin. The unit name is “kelvin” and the symbol is K, with no degree sign. Room temperature is about 293 K; a 37 °C body is 310.15 K; the surface of the sun is roughly 5,800 K.
Rankine, the Fahrenheit-sized absolute scale
Rankine is to Fahrenheit what Kelvin is to Celsius: an absolute scale with the same step size. °R = °F + 459.67, and 1 °R = 5/9 K. Zero Rankine is absolute zero, and 491.67 °R is the freezing point of water. It survives in some US engineering practice, especially steam tables and older HVAC work, because the degree size matches Fahrenheit instrumentation.
If a calculation mixes Rankine with Kelvin, convert both to Kelvin first. Mixing an absolute scale with a relative one is a reliable way to produce an answer that is wrong by hundreds of degrees.
Differences are not temperatures
This is the error that survives review because the arithmetic looks right. A temperature is a point on a scale; a temperature difference is an interval. The conversion formulas above include an offset, so applying them to a difference double-counts that offset. The correct rules are simpler: Δ°F = Δ°C × 1.8 and ΔK = Δ°C.
If an oven is set 25 °C hotter, that is 45 °F hotter, not 77 °F hotter. If tomorrow’s high is 5 °C above today’s, it is 9 °F above, not 41 °F. If a thermostat tolerance is ±2 K, it is ±2 °C and ±3.6 °F. Weather forecasts, tolerance bands, and setpoints all involve differences, and the offset formula is the wrong tool for them.
A related trap is averaging. Average the readings in one scale, then convert the average; do not convert each reading and average the results, because rounding at each step can shift the mean. The two orders agree exactly only when no rounding occurs.
Rounding by context
Weather is reported in whole degrees, so 22.2 °C is 22 °C and 71.6 °F is 72 °F. Oven dials move in 25 °F or 5 °C steps, so 176.7 °C becomes 175 °C or 180 °C depending on the recipe’s convention. Body temperature is usually quoted to one decimal: 37.0 °C or 98.6 °F. Laboratory work may need 0.01 K or better, and industrial specifications should state the tolerance explicitly.
Note that 98.6 °F is itself a rounded conversion of 37 °C, not an independently measured normal value. Converting it back gives 37.0 °C, and the apparent precision is an artifact of the rounding convention, not new information about human bodies.
Worked examples
Oven setting. 350 °F: (350 − 32) × 5/9 = 176.7 °C, which recipes round to 175 °C or 180 °C. The reverse: 180 °C × 1.8 + 32 = 356 °F, dialed as 350 °F.
Weather. 72 °F: (72 − 32) × 5/9 = 22.2 °C. A 101 °F fever: (101 − 32) × 5/9 = 38.3 °C.
Cold. −5 °F: (−5 − 32) × 5/9 = −20.6 °C. At −40, both scales read −40.
Laboratory. 298.15 K is 25.00 °C and 77.00 °F, a standard reference temperature for thermodynamic data.
High temperature. 1,500 °C is 2,732 °F and 1,773.15 K. At this range, rounding to the nearest degree is usually adequate, but the scale must still be stated.
Where temperature conversions go wrong
Multiplying before subtracting. 1.8 × 100 = 180 is not the Fahrenheit equivalent of 100 °C; the correct answer is 212 °F. The offset must be applied at the right point.
Adding 32 twice. Converting Celsius to Fahrenheit adds 32 once, after scaling. Converting back subtracts it once, before scaling.
Using 273 instead of 273.15. The difference is 0.15 K, which is negligible for weather and potentially significant in calibration or cryogenics.
Assuming 0 °F is absolute zero. It is not; absolute zero is −459.67 °F. The Fahrenheit zero was set to the temperature of a brine mixture, not to the bottom of the thermodynamic scale.
Treating heat index and wind chill as converted temperatures. They are calculated apparent temperatures, not unit conversions, and they apply only to human perception under specific conditions.
Frequently asked questions
What is the formula to convert Celsius to Fahrenheit?
°F = °C × 9/5 + 32. For 20 °C: 20 × 1.8 = 36, then + 32 = 68 °F.
What is the formula to convert Fahrenheit to Celsius?
°C = (°F − 32) × 5/9. For 68 °F: 36 × 5/9 = 20 °C.
How do I convert Celsius to Kelvin?
Add 273.15: K = °C + 273.15. So 0 °C = 273.15 K and 100 °C = 373.15 K.
Is a Celsius degree the same size as a Kelvin?
Yes. The kelvin and the degree Celsius have the same magnitude; only their zero points differ.
Why is −40 the same in both scales?
Because the 9/5 scaling and the 32-degree offset intersect there: −40 °C × 1.8 + 32 = −40 °F.
Can temperature be below 0 K?
No. Absolute zero is the minimum thermodynamic temperature. Some systems described as having “negative temperature” are hotter than any positive temperature in a specific statistical sense, not colder than 0 K.
Further reading
Keep exploring
Decision path