Coefficient of Variation Calculator
Calculate the coefficient of variation (CV) — the ratio of the standard deviation to the mean — as a decimal and a percentage. Paste raw data or enter your SD and mean, and optionally compare two datasets to see which is more consistent, whatever their units.
Textbook-accurate formula · Reviewed July 2026 · Aligned with standard coefficient of variation conventions
Enter Your Data
Works with commas, spaces, tabs, or line breaks. Decimals are fine.
Your Result
Variability Level
<10%Low
10–20%Moderate
20–35%High
>35%
What Is the Coefficient of Variation?
The coefficient of variation (CV) is a standardized measure of how spread out data is relative to its mean. It’s calculated by dividing the standard deviation by the mean, which strips away the units and leaves a pure ratio. That’s what makes CV so powerful: because it’s dimensionless, you can use it to compare the variability of datasets that have completely different units or wildly different averages — something the raw standard deviation can never do on its own.
A low coefficient of variation means the data clusters tightly around the mean, signalling consistency and stability. A high CV means the values are widely dispersed. CV is expressed either as a decimal (0.15) or, more commonly for reporting, as a percentage (15%). It’s also known as the relative standard deviation (RSD) when written as a percentage — the two are the same measure in different clothing.
The Coefficient of Variation Formula
The formula is simple and comes in two equivalent forms:
CV = s / x̄ (decimal) · CV = ( s / x̄ ) × 100% (percentage)
Here s is the standard deviation and x̄ is the mean. To get the decimal form, divide the standard deviation by the mean. To express it as a percentage, multiply by 100. That percentage form is identical to the relative standard deviation. For populations, the same formula uses the population standard deviation σ and population mean μ.
How to calculate the coefficient of variation step by step
- Calculate the mean of your dataset — add all values and divide by the count.
- Calculate the standard deviation — use the sample formula (÷ n − 1) for a subset of data, or the population formula (÷ N) for a complete set.
- Divide the standard deviation by the mean to get the CV as a decimal.
- Multiply by 100 if you want the percentage form.
The calculator above does every step automatically and shows the working. If you already have your standard deviation and mean, switch to “SD & mean” mode to jump straight to the answer.
How to Interpret the Coefficient of Variation
Since CV is a relative measure, a lower value always indicates less variability relative to the mean. There’s no single universal cutoff for “high” or “low” — it depends entirely on your field — but these general bands are a helpful starting point:
| CV Value | Interpretation | What it suggests |
|---|---|---|
| Below 10% | Very low variability | Highly consistent, stable data |
| 10% – 20% | Low variability | Reasonably consistent |
| 20% – 35% | Moderate variability | Noticeable spread around the mean |
| Above 35% | High variability | Widely dispersed; interpret with care |
These thresholds are rules of thumb, not hard rules. In analytical chemistry a CV above 2% might be a concern, while in finance or biology a CV of 30% could be completely normal. Always judge CV against the norms and requirements of your specific field.
Why Use CV Instead of Standard Deviation?
Imagine comparing the consistency of two very different things — say, the daily price swings of a $5 stock versus a $500 stock. The expensive stock will almost always have a larger standard deviation in dollar terms, but that doesn’t mean it’s more volatile in relative terms. Dividing by the mean levels the playing field. This is the classic reason to reach for CV: comparing variability across datasets with different scales or units. The comparison feature in the calculator above does exactly this — enter two datasets and it tells you which is more consistent, regardless of their units.
| Standard Deviation | Coefficient of Variation | |
|---|---|---|
| Units | Same as the data | Dimensionless (ratio or %) |
| Best for | Spread within one dataset | Comparing across datasets |
| Depends on scale? | Yes | No |
| Needs non-zero mean? | No | Yes |
To calculate the standard deviation on its own, use our Standard Deviation Calculator. For the same measure reported specifically as a percentage, see the Relative Standard Deviation Calculator. And for a complete breakdown of a dataset, try the Descriptive Statistics Calculator.
When not to use the coefficient of variation
CV only makes sense for ratio-scale data with a true, meaningful zero and a positive mean. It breaks down in two situations: when the mean is zero (the formula divides by zero) or very close to zero (tiny mean changes cause huge CV swings), and on interval scales with arbitrary zeros — temperature in Celsius or Fahrenheit is the classic example, where the same temperatures give different CVs depending on the scale. In those cases, CV can mislead rather than clarify.
Frequently Asked Questions
The coefficient of variation (CV) is the ratio of the standard deviation to the mean, calculated as CV = s / x̄. It measures how much variability exists relative to the average, and because it’s dimensionless, it lets you compare the spread of datasets with different units or magnitudes. It’s expressed as a decimal (0.15) or a percentage (15%), and a lower CV means more consistent data.
Yes, they’re the same measure. The coefficient of variation is usually written as a decimal (e.g., 0.15), while the relative standard deviation (RSD) is the same value written as a percentage (e.g., 15%). RSD also always uses the absolute value of the mean so it stays positive. In practice, RSD = |CV| × 100, so the two are interchangeable in most contexts.
Lower is generally better, since it means less relative variability, but “good” depends on the field. As a broad guide, a CV below 10% is very low, 10–20% is low, 20–35% is moderate, and above 35% is high. However, analytical chemistry often demands a CV under 2%, while a CV of 30% may be perfectly acceptable in biology or finance. Always compare against your field’s norms rather than a universal cutoff.
Calculate the CV of each dataset, then compare the values directly — the dataset with the lower CV has less variability relative to its mean, meaning it’s more consistent. Because CV is unitless, this works even when the datasets use different units or have very different averages. Turn on “Compare two datasets” in the calculator above to do this automatically and see which one wins.
CV can technically be negative if the mean is negative, since the standard deviation is always positive but the mean can be either sign — this is one small difference from RSD, which uses the absolute mean and stays positive. CV can also exceed 100% (or 1.0 as a decimal) when the standard deviation is larger than the mean, which happens with highly variable data such as some income or waiting-time distributions.
Avoid CV when your mean is zero or near zero, because dividing by it produces undefined or wildly unstable results. Also avoid it for interval-scale data with an arbitrary zero point, such as temperature in Celsius or Fahrenheit, where the same data gives different CVs on different scales. CV is only meaningful for ratio-scale data with a true zero and a clearly non-zero mean.
Methodology & formulas used
This calculator uses standard textbook formulas, reviewed July 2026. All computation runs locally in your browser; your data is never uploaded.
- Mean: x̄ = Σxᵢ / n
- Sample standard deviation: s = √[ Σ(xᵢ − x̄)² / (n − 1) ]
- Population standard deviation: σ = √[ Σ(xᵢ − x̄)² / N ]
- Coefficient of variation (decimal): CV = s / x̄
- Coefficient of variation (percent): CV% = (s / x̄) × 100
The variability bands (very low <10%, low 10–20%, moderate 20–35%, high >35%) are general reference points shown using the absolute value of CV; specific fields apply their own thresholds. When comparing two datasets, the one with the smaller absolute CV is the more consistent. CV is undefined when the mean is zero. Results are rounded for display but computed at full precision.
References
- Wikipedia. Coefficient of Variation — definition, formula, and cautions. en.wikipedia.org
- NIST/SEMATECH. Measures of Scale — e-Handbook of Statistical Methods. itl.nist.gov
- Statistics How To. Coefficient of Variation — Definition and Examples. statisticshowto.com