What Is Relative Standard Deviation?
Relative standard deviation, also written %RSD, shows how large the spread of your data is compared with its average. A standard deviation of 2 means something different when the mean is 10 than when it is 1,000; dividing by the mean puts both on the same scale so you can compare them.
RSD is identical to the coefficient of variation (CV) when expressed as a percentage, and many labs report it as %RSD or %CV for repeatability and precision. Choose sample or population standard deviation in the calculator, and read the plain-English verdict beneath the results for guidance on how to interpret the number.
Below you will find the RSD formula, a variables table, a worked example, when RSD is misleading and answers to common questions such as RSD versus standard deviation.
How to Use the Relative Standard Deviation Calculator
- Enter your repeated measurements, separated by commas, spaces or new lines.
- Choose sample (the usual choice for replicate measurements) or population.
- Read the %RSD. A lower value means the measurements are more tightly grouped relative to their mean.
Relative Standard Deviation Formula
Relative standard deviation
Coefficient of variation
Relative Standard Deviation Variables Explained
| Symbol | Meaning | Notes |
|---|---|---|
| RSD | Relative standard deviation | Usually quoted as %RSD |
| s | Standard deviation | Sample (n − 1) or population (n) |
| x̄ | Mean of the data | Should not be close to zero |
| n | Number of measurements | Needs at least two for a sample SD |
Relative Standard Deviation Worked Example
Using the example values (press “Load example” above to try them):
- Mean: x̄ = 50.4 ÷ 5 = 10.08.
- Sum of squared deviations: Σ(x − x̄)² = 0.228.
- Standard deviation: s = √(Σ(x − x̄)² ÷ (n − 1)) = 0.238747.
- Relative standard deviation: RSD = (SD ÷ |x̄|) × 100 = (0.238747 ÷ 10.08) × 100 = 2.3685 %.
Result: Relative standard deviation (RSD) = 2.3685 %; Sample standard deviation (s) = 0.238747.
Interpreting %RSD
%RSD expresses the standard deviation as a share of the mean, so you can compare precision between measurements of very different size. A 0.5 % RSD on a 100 mg/L standard and a 0.5 % RSD on a 1 mg/L standard are equally precise in relative terms.
The precision rating in the results uses broad bands. Real acceptance limits come from your method, regulation or quality system, so always compare with your own criterion.
| Application | Commonly seen limit |
|---|---|
| Instrument injection repeatability (HPLC, area) | ≤ 1 to 2 % |
| Assay of a drug substance, replicate preparations | ≤ 2 % |
| Impurities at low level | ≤ 5 to 10 % |
| Trace analysis near quantitation limit | ≤ 15 to 20 % |
| Bioanalytical method validation (LLOQ) | ≤ 20 % |
RSD, CV and %CV: the same thing
Analytical chemists say %RSD, biologists and economists say CV or %CV, and finance talks about volatility relative to the mean. All are the standard deviation divided by the mean. The coefficient of variation is the plain ratio (0.05) and the RSD is that ratio times 100 (5 %).
When RSD is misleading
- Data on an interval scale with an arbitrary zero (°C, °F, pH) gives meaningless RSD values because the mean depends on where zero is placed.
- A mean close to zero makes the RSD explode even when the spread is small.
- Skewed or log-normal data are better described by a geometric CV.
- Very few replicates make the SD, and therefore the RSD, uncertain; three replicates give a very rough estimate.
Reporting RSD in the lab
- Report the number of replicates (n) with the %RSD.
- State whether you used the sample (n − 1) SD, which is the usual choice.
- Quote the RSD to a sensible number of figures, usually one or two decimals.
- If you report the RSD of an average of replicates, use the SEM divided by the mean (shown above).
Where Relative Standard Deviation Is Used
Laboratory precision
Summarising the repeatability of replicate measurements as a single percentage.
Comparing variability across scales
Judging which of two data sets with different means is relatively more spread out.
Quality control
Tracking how consistent a process output is from batch to batch.
Reporting results
Giving the reader a unit-free measure of spread alongside the mean.
Common Relative Standard Deviation Mistakes to Avoid
- RSD only makes sense for ratio-scale data with a true zero, such as weights or concentrations. It is not meaningful for temperatures in °C.
- If the mean is close to zero the RSD becomes very large and unstable.
- Acceptable RSD limits depend on the method and the field, so compare your result with the requirement for your own procedure.
Relative Standard Deviation Terms and Definitions
- Precision
- How closely repeated measurements agree with each other.
- Accuracy
- How close the measurements are to the true value; RSD does not measure this.
- Replicate
- A repeated measurement of the same sample under the same conditions.
- Repeatability
- Precision under the same conditions over a short time.
- Coefficient of variation
- Standard deviation divided by the mean.
Relative Standard Deviation FAQs
How do I calculate %RSD in Excel?
Use =STDEV.S(range)/AVERAGE(range)*100 for a sample, or STDEV.P for a population.
What is an acceptable %RSD?
It depends on the method and regulatory guidance. Pharmaceutical assays often require ≤ 2 %, while trace methods may allow 15 to 20 %.
What is the difference between RSD and standard error?
RSD describes the spread of individual measurements relative to their mean. The standard error describes the uncertainty of the mean itself and shrinks with √n.
Why is RSD not valid for temperature in Celsius?
Celsius has an arbitrary zero, so the mean (and therefore the ratio) changes when you change the scale. Use Kelvin or choose a different measure.
Should I use sample or population SD for RSD?
Use the sample SD (n − 1) for replicate measurements, which are a sample of all possible measurements.
Can RSD be above 100 %?
Yes, when the standard deviation is larger than the mean, which is common for skewed data such as counts with many small values and a few huge ones.
What is the difference between RSD and standard deviation?
Standard deviation is in the units of your data. RSD divides it by the mean and expresses it as a percentage, which lets you compare variability between data sets of different scales.
Is RSD the same as the coefficient of variation?
Yes, they are the same quantity. RSD is the percentage form (CV × 100) and is common in analytical chemistry.
What is a good RSD value?
Lower is more precise, but what counts as acceptable depends on the method, concentration level and regulatory or lab requirements.
Can RSD be negative?
It is reported as a positive value because the calculation uses the absolute value of the mean.