How do you calculate relative standard deviation (RSD)?

Divide the sample standard deviation by the mean and multiply by 100: RSD (%) = (s / x̄) × 100. For example, a standard deviation of 2 with a mean of 40 gives an RSD of 5%. You can get %RSD automatically with our RSD calculator.

What is a good RSD value?

It depends on the application. In analytical chemistry an RSD below about 2% is typically excellent for HPLC system suitability, while 5–15% may be acceptable at low concentrations. Regulatory methods set explicit limits, so always check the acceptance criterion for your specific assay.

Is relative standard deviation the same as coefficient of variation?

Yes. RSD and coefficient of variation (CV) are mathematically identical—both express the standard deviation as a proportion of the mean. RSD is usually written as a percentage (e.g., 5.2%) while CV is sometimes left as a decimal (0.052). Laboratories tend to prefer the term RSD.

Can RSD be greater than 100%?

Yes. When the standard deviation exceeds the mean, RSD rises above 100%, which signals very high relative variability. This commonly happens with data near zero or with highly skewed distributions. RSD is only meaningful for ratio-scale data with a positive, non-zero mean.

Relative Standard Deviation (RSD & %RSD): Formula & Guide

What is Relative Standard Deviation?

Relative Standard Deviation (RSD), also known as coefficient of variation (CV), is a standardized measure of dispersion that expresses the standard deviation as a percentage of the mean. It's the gold standard for assessing precision in analytical chemistry, pharmaceutical testing, and quality control laboratories.

Unlike absolute standard deviation, RSD allows you to compare variability across measurements with different scales or units. A standard deviation of 5 mg/L might be excellent for one analysis but unacceptable for another—RSD puts everything on a common scale. You can compute it instantly from raw data or from a known mean and SD with our free relative standard deviation calculator.

RSD vs CV

RSD and coefficient of variation (CV) are mathematically identical. RSD is typically expressed as a percentage (e.g., 5.2%), while CV may be expressed as a decimal (0.052). In laboratory settings, RSD is the more common terminology.

RSD Formula and Calculation

Relative Standard Deviation

RSD (%) = (s / x̄) × 100

Where s is the sample standard deviation and x̄ is the sample mean. The calculation is straightforward:

Calculate the Mean

Sum all values and divide by the number of measurements.

Calculate Standard Deviation

Find the square root of the variance (sum of squared deviations from mean, divided by n-1).

Divide and Multiply

Divide SD by mean, then multiply by 100 to express as percentage.

python

import numpy as np

def calculate_rsd(data):
    """Calculate Relative Standard Deviation"""
    mean = np.mean(data)
    std = np.std(data, ddof=1)  # Sample SD with Bessel's correction
    rsd = (std / mean) * 100
    return rsd

# Example: Analytical measurements
measurements = [98.5, 101.2, 99.8, 100.5, 99.1]
rsd = calculate_rsd(measurements)
print(f"RSD = {rsd:.2f}%")  # Output: RSD = 1.11%

Interpreting RSD Values

The acceptable RSD depends on your application, concentration levels, and regulatory requirements:

RSD < 2%:Excellent precision; typical for well-validated HPLC assays and reference standards
RSD 2-5%:Good precision; acceptable for most pharmaceutical content uniformity tests
RSD 5-10%:Moderate precision; may be acceptable for biological assays or trace analysis
RSD 10-15%:Higher variability; typical for immunoassays and bioanalytical methods
RSD > 15%:Poor precision; may indicate method problems or sample inhomogeneity

Concentration Matters

RSD typically increases at lower concentrations due to greater relative impact of measurement uncertainty. The Horwitz equation predicts this relationship: RSD doubles for every 10-fold decrease in analyte concentration.

Regulatory Requirements

Regulatory agencies set specific RSD requirements for different test types:

FDA/ICH Guidelines

System suitability: RSD ≤ 2% (5 injections) · Method precision: RSD ≤ 2% typically · Content uniformity: RSD requirements in USP <905> · Dissolution: RSD ≤ 20% at early timepoints

Bioanalytical Methods

QC samples: RSD ≤ 15% (≤20% at LLOQ) · Calibrators: At least 75% within ±15% · Incurred sample reanalysis: 67% within 20%

Laboratory Applications

RSD is essential across analytical sciences:

Method Validation:Demonstrating precision, repeatability, and intermediate precision during method development
System Suitability:Daily verification that HPLC systems are performing within specifications
Stability Studies:Monitoring analytical precision over long-term stability programs
Method Transfer:Comparing precision between laboratories or instruments
Quality Control:Batch-to-batch consistency in manufacturing and release testing

Worked Examples

Example 1: HPLC System Suitability

Five replicate injections give peak areas: 1,245,678; 1,251,234; 1,248,901; 1,244,567; 1,249,890 Mean = 1,248,054 | SD = 2,689 | RSD = 0.22% - Passes ≤2% criterion

Example 2: Content Uniformity

Ten tablet assays: 99.2%, 101.5%, 98.8%, 100.3%, 99.7%, 100.8%, 99.1%, 101.2%, 100.1%, 99.5% Mean = 100.02% | SD = 0.91% | RSD = 0.91% - Excellent uniformity

Calculate RSD automatically

Skip the manual arithmetic: paste your replicate measurements into the RSD calculator to get %RSD, the coefficient of variation, and step-by-step working in one click.

Sources

References and further authoritative reading used in preparing this article.

← Learning Center