Written by Dackohn; last reviewed November 2026.

The coefficient of variation is the standard deviation divided by the mean, usually as a percentage. It compares relative variability across different units: heights with CV 4.7% vary less, relatively, than weights with CV 17.1%. It only works for data with a true zero and means well above zero.

The formula

CV = (s / x̄) × 100%

A CV of 20% means the SD is one-fifth of the mean. Because the units cancel, the CV has no unit, and that is what makes comparisons possible.

Worked example

Are people more variable in height or in weight?

VariableValuesMeanSDCV
Height (cm)170, 165, 180, 175, 1601707.914.7%
Weight (kg)60, 72, 85, 68, 556811.6017.1%

Comparing 7.91 cm with 11.60 kg is meaningless, but the CVs show weight varies far more relative to its size.

Where it is useful

When not to use it

Without a true zero. Temperatures of 10, 20 and 30 °C have a mean of 20 and SD of 10, so the CV is 50%. The same temperatures in Fahrenheit (50, 68, 86) give 26.5%. Nothing physical changed, only the scale, so the result is meaningless. Use the CV only for ratio data.

With means near zero. Dividing by a tiny mean makes the CV explode. This often happens with changes or differences, which can be positive or negative.

With small samples. The CV is itself uncertain and slightly biased. A common correction multiplies by (1 + 1 / (4n)).

What is a good CV?

It depends entirely on the field. A clinical assay might require under 10%, while a CV of 50% is unremarkable for incomes. Compare with similar measurements in your field.

In Excel

=STDEV.S(A1:A5)/AVERAGE(A1:A5), formatted as a percentage.

Try it yourself

Coefficient of Variation Calculator →Get the CV, or compare two datasets. Descriptive Statistics Calculator →Mean, SD and more from your data.

Frequently asked questions

What does the coefficient of variation tell you?

How large the standard deviation is relative to the mean, which allows comparisons across different scales and units.

Can the CV be greater than 100%?

Yes, when the SD exceeds the mean, as often happens with strongly right-skewed data.

What is the difference between CV and standard deviation?

The SD is in the original units. The CV is unit-free, so it can compare different variables.

Related reading: Types of variables · How to calculate standard deviation · Skewness and kurtosis


Sources & further reading: NIST/SEMATECH e-Handbook of Statistical Methods · OpenStax Introductory Statistics. How we compute & validate: methodology. Written & reviewed by Dackohn; last reviewed August 2026.