📐 Calibration Curve Calculator
Fit a linear calibration curve from your standards, then use it to calculate the concentration of an unknown sample.
| x (µg/mL) | y (signal) |
|---|---|
| 0 | 52 |
| 2 | 5120 |
| 4 | 10080 |
| 6 | 15200 |
| 8 | 20150 |
Frequently asked questions
How many calibration points do I need?
A minimum of 3 is required to calculate a line, but analytical guidelines (e.g. ICH Q2(R1)) typically recommend at least 5-6 standard concentrations spanning the expected range of unknown samples, ideally with each level run in replicate.
What does R² tell me about my calibration curve?
R² (coefficient of determination) indicates how well the linear model fits your data, from 0 to 1. Values above 0.995 are generally expected for a reliable calibration curve, but R² alone doesn't guarantee accuracy — always check that your unknown sample's signal falls within the calibrated range, not extrapolated beyond it.
Can I use this for non-linear calibration curves?
No — this tool fits a straight line (simple linear regression) only. If your response is not linear across the range of interest (common at high concentrations), you'll need a different regression model or should narrow your calibration range.
Is it safe to extrapolate beyond my highest or lowest standard?
No — quantifying a sample whose signal falls outside the calibrated range (extrapolation) is not recommended, since you have no evidence the linear relationship holds there. Dilute or concentrate the unknown sample to bring its signal within the calibrated range instead.
Accuracy & how this is derived
Derivation: Standard least-squares linear regression (y = mx + b, minimizing sum of squared residuals) with R² computed as 1 minus the ratio of residual to total sum of squares. Unknown concentration is found by algebraic inversion of the fitted line.
Validated against: Standard ordinary least-squares regression, consistent with ICH Q2(R1) guidance on calibration curve linearity for analytical method validation.
⚠️ For educational and research support only — verify critical results independently before use in regulated, clinical, or publication-bound work.
✔ Last updated: July 2026 · Report an error
Building and using a calibration curve
A calibration curve is built by measuring the instrument response (signal) for a series of standards of known concentration, then fitting a linear regression line through the points using the least-squares method. Once the slope and intercept are known, the concentration of an unknown sample can be calculated directly from its measured signal by algebraically inverting the fitted equation, provided the unknown's signal falls within the range spanned by the calibration standards.