How do I convert Celsius to Fahrenheit?

Multiply by 9/5 (1.8) and add 32. So 20 °C is 20 × 1.8 + 32 = 68 °F. To go back, subtract 32 and multiply by 5/9.

Why is an uncertainty of 0.5 °C not 32.9 °F?

Because 0.5 °C is a size, not a position on the scale. A size converts by the factor 1.8 alone, giving 0.9 °F. Adding 32 is correct only for a temperature such as 0.5 °C, which is 32.9 °F, and getting this wrong in a certificate understates or overstates the uncertainty by a large amount.

At what temperature are Celsius and Fahrenheit equal?

−40. At that point °C and °F show the same number, which is also a handy check on any converter.

What is the Rankine scale?

An absolute scale with Fahrenheit-sized degrees: 0 °R is absolute zero, so °R = °F + 459.67. It is used in some US thermodynamics work.

Temperature converter: °C, °F, K and °R, points and intervals

Convert a temperature between Celsius, Fahrenheit, kelvin and Rankine, and convert a temperature difference or uncertainty separately, because the two do not follow the same rule.

A temperature

ValueScale

Celsius: 20 °C
Fahrenheit: 68 °F
kelvin: 293.15 K
Rankine: 527.67 °R

Runs entirely in your browser · nothing you enter is sent to a server

A temperature difference or an uncertainty

Interval (±)Scale

Celsius: 0.5 °C
Fahrenheit: 0.9 °F
kelvin: 0.5 K
Rankine: 0.9 °R

An interval has no offset: a change of 1 °C is a change of 1 K and 1.8 °F. Use this box for tolerances, drift, resolution and uncertainties, and the box above only for an actual temperature.

How the calculation works

01A temperature point converts with an offset: K = °C + 273.15, °F = °C × 9/5 + 32, °R = K × 9/5. Every conversion here passes through kelvin, so no pair of scales can disagree with another.
02A temperature interval converts with the scale factor alone: one kelvin is one degree Celsius is 1.8 degrees Fahrenheit or Rankine. This is the rule for tolerances, drift, resolution and expanded uncertainties.
03A value below absolute zero (0 K, −273.15 °C, −459.67 °F) is refused rather than converted, because no physical temperature exists there and a typo should not produce an answer.

Limitations

Conversion is exact only to the digits shown. The kelvin–Celsius offset is exactly 273.15 by definition.
It converts a value you already trust. It does not know whether a reading is a temperature or an interval, so choose the box that matches.

Frequently asked questions

How do I convert Celsius to Fahrenheit?: Multiply by 9/5 (1.8) and add 32. So 20 °C is 20 × 1.8 + 32 = 68 °F. To go back, subtract 32 and multiply by 5/9.
Why is an uncertainty of 0.5 °C not 32.9 °F?: Because 0.5 °C is a size, not a position on the scale. A size converts by the factor 1.8 alone, giving 0.9 °F. Adding 32 is correct only for a temperature such as 0.5 °C, which is 32.9 °F, and getting this wrong in a certificate understates or overstates the uncertainty by a large amount.
At what temperature are Celsius and Fahrenheit equal?: −40. At that point °C and °F show the same number, which is also a handy check on any converter.
What is the Rankine scale?: An absolute scale with Fahrenheit-sized degrees: 0 °R is absolute zero, so °R = °F + 459.67. It is used in some US thermodynamics work.

Need the reasoning, not just the number?

Ask ValiTrac AI and see the standards evidence and the engine calculation behind the answer.

Ask ValiTrac AI

Equipment

Thermocouple types and tolerance classes (IEC 60584)

Types K, J, T, N and E each have a defined EMF curve and tolerance classes 1, 2 and 3. The tolerance applies to new wire, not to a year-old thermocouple.

2026-09-13 · 4 min read

Equipment

Pt100 tolerance classes explained (IEC 60751)

Classes AA, A, B and C widen with temperature: Class A is ±0.15 °C at 0 °C but ±0.55 °C at 200 °C. A tolerance class is not a measurement uncertainty.

2026-09-13 · 3 min read

Calibration