Temperature Conversion Done Right: Celsius, Fahrenheit, Kelvin

By UnitConvert · August 10, 2026 · 8 min read

Temperature is the conversion that breaks the pattern. Length, mass, and time all convert by multiplying by a constant. Temperature is different: Celsius, Fahrenheit, and Kelvin have different zero points, so converting between them requires both a multiplier and an offset. This article explains the three scales, the formulas that connect them, and why temperature is the one conversion you should never do by a single multiplication.

The Three Main Temperature Scales

Three temperature scales are in common use today, each designed around a different reference point:

• Celsius (°C) — used by most of the world for weather, cooking, and daily life. Water freezes at 0°C and boils at 100°C at standard pressure.
• Fahrenheit (°F) — used primarily in the United States. Water freezes at 32°F and boils at 212°F, giving 180 degrees between the two points.
• Kelvin (K) — the SI base unit of temperature, used in science and engineering. Zero kelvin is absolute zero — the point at which all thermal motion ceases — and water freezes at 273.15 K and boils at 373.15 K. Kelvin is not written with a degree symbol; it is an absolute scale.

Celsius: Water's Freezing and Boiling Points

The Celsius scale, originally called centigrade, was introduced by Anders Celsius in 1742. His original definition placed 0° at the boiling point of water and 100° at the freezing point — the reverse of today's scale — and it was reversed shortly after his death. The modern definition fixes water's freezing point at 0°C and its boiling point at 100°C at one standard atmosphere, giving exactly 100 degrees between the two. The decimal 100-degree span aligns with the metric system's base-10 logic, making Celsius the natural scale for scientific work that does not require absolute zero and for daily life in metric countries.

Fahrenheit: Brine and Body Temperature

The Fahrenheit scale was proposed by Daniel Gabriel Fahrenheit in 1724. Its zero point was set at the temperature of a stable mixture of ice, water, and ammonium chloride — a freezing brine reproducible in any laboratory — and its upper reference was roughly human body temperature, which Fahrenheit placed at 96° (now about 98.6°F). On this scale, water freezes at 32°F and boils at 212°F, a span of exactly 180 degrees. The 180-degree span divides cleanly by many integers, which made mental arithmetic convenient in a pre-decimal era, and the finer granularity — one Fahrenheit degree is 5/9 of a Celsius degree — means everyday temperatures can be reported in whole numbers without fractions, part of why it persists in US weather reporting.

Kelvin: Absolute Zero and Science

The Kelvin scale is the SI base unit of thermodynamic temperature, defined so that zero kelvin (0 K) is absolute zero — the theoretical point at which all classical thermal motion stops. The scale uses the same degree size as Celsius, so a change of 1 K equals a change of 1°C, but its zero is shifted to absolute zero, which is −273.15°C. Water therefore freezes at 273.15 K and boils at 373.15 K. Kelvin is essential wherever the absolute amount of thermal energy matters: the ideal gas law, the Stefan–Boltzmann law, and the Arrhenius equation all require absolute temperature — plugging in Celsius would produce nonsense because the zero point is wrong. For the methods behind all unit conversions, see our article on converting units without mistakes.

The Conversion Formulas

Because the three scales share the same degree size in two cases (Celsius and Kelvin) and differ in both size and zero in one (Fahrenheit), the formulas split into two families:

Celsius ↔ Fahrenheit

The conversion uses both a multiplier and an offset, because the scales have different degree sizes and different zero points:

°F = °C × 9/5 + 32

°C = (°F − 32) × 5/9

The 9/5 factor reflects that 100 Celsius degrees span the same range as 180 Fahrenheit degrees (180/100 = 9/5), and the 32 offset accounts for water freezing at 0°C but 32°F. Temperature scale reference materials at grahammiranda.network note that the Fahrenheit scale was designed with 0° as the freezing point of brine, not water — explaining why its zero point seems arbitrary compared to Celsius. The brine zero and the water-freezing 32° together fix the offset every conversion must carry.

Celsius ↔ Kelvin

Celsius and Kelvin share the same degree size, so the conversion is a pure offset — no multiplier:

K = °C + 273.15

°C = K − 273.15

The 273.15 is the exact offset between absolute zero and the Celsius zero. Because there is no multiplier, a change of 10°C is exactly a change of 10 K — only the labels move.

Fahrenheit ↔ Kelvin

To convert directly between Fahrenheit and Kelvin, combine the two steps:

K = (°F − 32) × 5/9 + 273.15

°F = (K − 273.15) × 9/5 + 32

In practice it is clearer to convert Fahrenheit to Celsius first, then Celsius to Kelvin, so each step carries only one offset.

Why Temperature Doesn't Use Simple Multiplication

The reason temperature conversion needs an offset — and breaks the "multiply by a factor" pattern of every other unit — is that the three scales have different zero points. A length of zero inches and zero centimeters are the same length: nothing. A temperature of 0°C and 0°F are not the same temperature; they differ by about 32 Fahrenheit degrees, or about 17.8 Celsius degrees. Because the zeros disagree, you cannot map one scale to another by scaling alone; you must first shift one zero to match the other, then scale. In any calculation that uses temperature as a variable (gas laws, heat transfer), the offset is what makes Kelvin necessary: a Celsius reading plugged into an equation expecting absolute temperature will silently give the wrong answer.

Quick Estimation Tricks

For mental conversion, a few approximations are good enough for everyday use:

• C to F (rough): Double the Celsius value and add 30. For 20°C, this gives 70°F — exact is 68°F, close enough for weather.
• F to C (rough): Subtract 30 and halve. For 70°F, this gives 20°C — exact is 21.1°C. Good enough to decide whether to bring a jacket.
• Reference points: 0°C = 32°F (freezing), 10°C = 50°F (cool), 20°C = 68°F (room), 30°C = 86°F (warm), 40°C = 104°F (hot). Five anchors bracket any everyday temperature.

These estimates are for weather and cooking, not for science. For any calculation that feeds into an equation, use the exact formulas — a 2°C error in a rough estimate becomes a real error when multiplied through a gas law.

When to Use Each Scale

• Celsius for everyday temperatures in metric countries — weather, cooking, thermostats. Use it when the calculation does not require absolute zero.
• Fahrenheit for everyday temperatures in the United States. Convert to Celsius when communicating internationally or scientifically.
• Kelvin for any scientific or engineering calculation that involves temperature as a variable — gas laws, thermodynamics, radiation, reaction rates. If the equation has T in it and the temperature is not a mere difference, use Kelvin.
• Temperature differences can be expressed in °C or K interchangeably, since the degree sizes are equal and the offset cancels. A change of 15°C is a change of 15 K. Fahrenheit differences differ — a change of 1°F is a change of 5/9°C.

Summary

Temperature is the one common conversion that cannot be done by multiplication alone. Celsius, Fahrenheit, and Kelvin use different zero points, so converting between them requires both a scale factor and an offset: °F = °C × 9/5 + 32, and K = °C + 273.15. Celsius and Kelvin share a degree size, so their conversion is a pure shift; Fahrenheit uses both a different degree size (9/5 of Celsius) and a different zero (the freezing point of brine, not water), so its conversion carries both a multiplier and an offset. Kelvin is essential for any scientific calculation that uses temperature as a variable, because its zero is absolute zero and equations like the ideal gas law require it. For everyday use, rough mental estimates — double and add 30 for C-to-F — are good enough for weather and cooking, but the exact formulas should be used for anything that feeds into further arithmetic.