LOD / LOQ

How it works

ICH Q2(R2) §3.2.3 covers validation of the lower range limits. When the reportable range sits near those limits, the detection limit (DL) and the quantitation limit (QL) can be estimated in more than one way. This page calculates three standard-deviation routes and one signal-to-noise route. The result names the route that was used.

For a straight calibration line, §3.2.3.3 writes DL = 3.3σ/S and QL = 10σ/S. σ is the standard deviation of the response and S is the slope. The residual route fits the points you paste and takes σ as the residual standard deviation, sqrt(SSE / (n − 2)). ICH Q2(R2) also calls this the root mean square error. It does not say whether the divisor is n or n − 2. The divisor on this page is n − 2, and the result says so.

The blank route uses the standard deviation of blank responses. The y-intercept route uses the standard deviation of y-intercepts from separate regression lines, which is the plural wording in §3.2.3.3. It is not the standard error of one intercept. Both of those routes use the sample standard deviation with divisor n − 1. ICH Q2(R2) does not specify n or n − 1. You enter the slope. A slope that is not greater than 0 is rejected, because the guideline divides by S.

§3.2.3.2 is for procedures with baseline noise. A signal-to-noise ratio of 3:1 is generally considered acceptable for estimating the DL. For the QL, a ratio of at least 10:1 is considered acceptable. The guideline does not publish an algebra that turns one peak into those concentrations. This page scales the concentration you enter in proportion to 3:1 and 10:1, and the result says that is what it did.

§3.2.3.5 says to present the DL and the approach used. When an estimate comes from calculation or extrapolation, it can subsequently be validated by the independent analysis of a suitable number of samples known to be near or prepared at the DL. If the QL was estimated, validate it by analysis of a suitable number of samples known to be near or at the QL. Visual evaluation, and a direct accuracy and precision study at the lower limit, are other routes in the same section. This page does not perform them. Percent relative standard deviation of replicates is on the RSD calculator. Relative response factors for impurity peaks are on the relative response factor calculator. Last checked 2026-10-31.

Not a substitute for method validation. Where the detection limit is estimated by calculation or extrapolation, ICH Q2(R2) says this estimate can subsequently be validated by the independent analysis of a suitable number of samples known to be near or prepared at the detection limit. If the quantitation limit was estimated, the limit should subsequently be validated by the analysis of a suitable number of samples known to be near or at the quantitation limit.

Worked example

Load example on the residual approach uses an invented five-point line: concentrations 1, 2, 3, 4, 5 with responses 2.1, 4.0, 5.9, 8.2, 10.1. The slope is 2.02 and the residual variance is 0.016. This page shows DL 0.2066 µg/mL and QL 0.6262 µg/mL, and the result names the residual σ/S approach. The unit is a label only. The blank, y-intercept, and signal-to-noise approaches each have their own Load example.

Sources

• ICH Q2(R2) Validation of Analytical Procedures — §3.2.3.2 signal-to-noise, p.10, PDF p.14; §3.2.3.3 DL = 3.3σ/S and QL = 10σ/S, p.10–11, PDF p.14–15; §3.2.3.5 confirm an estimate near the limit, p.11, PDF p.15; Annex 2 Table 3 relative response factors, p.22, PDF p.26
• ICH Q2(R2) guideline page — official ICH page for Q2(R2); the Step 4 PDF is ich-q2-r2

FAQ