Recovery and Bias Calculator
Enter the nominal (reference) and measured values to calculate the percent recovery, the systematic error (bias) and the precision statistics. For analytical method validation under ISO/IEC 17025 and EURACHEM guidance.
Acceptance criteria
| # | Sample ID (optional) | Nominal / reference value | Measured value | % Recovery | Absolute error | Relative error % |
|---|
For spike recovery, enter the amount added as the nominal value and the amount found (spiked minus unspiked result) as the measured value.
What is percent recovery in method validation?
In analytical chemistry, percent recovery measures what fraction of the analyte present in a sample is actually detected and quantified by the method. It is the usual way to assess trueness: you analyze a sample whose content is known (a certified reference material, a standard or a spiked sample) and compare the result with the expected value. A recovery of 100 % means the method introduces no bias: what is in the sample is exactly what you measure.
This is not the same as the percent recovery of a recrystallization or synthesis, where you compare the mass of product obtained with the mass you started with.
% Recovery = (measured value / nominal value) × 100
Absolute error = measured value − nominal value
Relative error (%) = (absolute error / nominal value) × 100
Bias (%) = mean % recovery − 100
The summary also reports the standard deviation of the recoveries and the coefficient of variation (CV%), which in pharmaceutical work is usually called the relative standard deviation (%RSD). Recovery tells you how close you are to the true value; CV% tells you how repeatable that recovery is.
Typical acceptance criteria
80 – 120 % → routine methods (water, food, industrial samples)
90 – 110 % → precision methods (pharmaceutical, accredited methods)
95 – 105 % → high-accuracy methods (pharmacopeial, reference methods)
CV% < 5 % → acceptable repeatability in most cases
These are typical values, not universal limits. The criteria that apply to your method come from your validation plan and from the guideline or standard it follows (for example AOAC, ICH Q2 or a pharmacopeia), which may set different limits depending on the analyte, the matrix and the concentration level. Enter your own limits in the fields above.
Frequently asked questions
Can I use this calculator for spike recovery?
Yes, as long as you subtract the unspiked result first. Enter the amount of analyte added as the nominal value and the amount found (spiked result minus unspiked result) as the measured value. The calculator does not take the unspiked sample as a separate input, so that subtraction is up to you, in the same units.
How many samples do I need to evaluate recovery?
EURACHEM recommends at least 6 independent determinations at each concentration level. For a full validation, at least 3 concentration levels are evaluated (low, medium and high within the working range). ICH Q2 sets a minimum of 9 determinations over at least 3 levels for accuracy.
Does a low recovery mean the method is bad?
Not necessarily. A consistently low recovery (e.g. 85 %) may be acceptable if it is constant and known; in that case a correction factor is applied. What matters is that the recovery is reproducible: a high CV% is a bigger problem than a recovery that is far from 100 % but stable.
Related tools
Measurement Uncertainty CalculatorCombine Type A and Type B uncertainty components and get the expanded uncertainty of your measurement.
Related tools
LOD and LOQ CalculatorCalculate the limit of detection and limit of quantitation of your method.
Related tools
Grubbs and Dixon Q Test CalculatorCheck your replicates for outliers before calculating the mean recovery.
This calculator is also available in Spanish.