Detection Limit and Quantification Limit Are Not Interchangeable

Overview

Yes — detecting a signal and quantifying it reliably are different tasks. A laboratory may report that an analyte was "detected" because the instrument recorded a signal above the background, but that same signal can be too small, noisy, or uncertain to support a precise concentration value with the accuracy and reliability required for decision-making. This article explains what limits of detection (LOD) and limits of quantification (LOQ or reporting limit) mean, how they relate to desired accuracy and working range, and why a detected analyte often cannot be reported as a precise quantity without additional evidence about method performance and context .

Why detection ≠ reliable quantification

An instrument can produce a measurable signal from a tiny amount of analyte. But measurement uncertainty grows as signal approaches background levels. When uncertainty is large relative to the measured value, the concentration estimate is not reliable for many uses. The difference between "present" and "present at X amount ± Y" is the key practical distinction .

LOD, LOQ and reporting limits — definitions and practical meaning

Key technical ideas behind those definitions

Why a detected analyte may not support a precise quantity

How laboratories and readers should treat low-level findings

A brief table of evidence types and what they tell you

| Evidence type | What it supports | |---|---| | Blank measurements, signal-to-noise data | Establishes LOD (presence vs absence) | | Replicate low-level standards, precision/accuracy data | Establishes LOQ (reliable quantification) | | Matrix spike/recovery and interference studies | Shows LOQ and bias control in specific matrices | | Calibration curve residuals and range | Shows where quantification model is valid |

What remains unresolved without method-specific data You cannot infer from a single detected signal that the lab's LOQ was met, that matrix effects were controlled, or that the reported number meets the precision and accuracy requirements for a particular decision. Determining whether a reported concentration is fit for purpose requires method-specific validation data and, when appropriate, current primary evidence for the exact matrix and use case . For general descriptions of analytical screening and reporting practices, see laboratory method summaries that document their LOD/LOQ approach and validation work .

In short: detection tells you an analyte likely exists above background; quantification requires additional, documented performance — precision, accuracy, calibration, and matrix validation — before a numeric concentration should be treated as reliable for decision-making .

• Detection asks: Is there a measurable signal that is distinguishable from blank or background noise?
• Quantification asks: What is the concentration, and how precise and accurate is that number?
• Limit of Detection (LOD): The smallest amount or concentration of analyte that produces a signal statistically distinguishable from the blank. LOD is tied to signal-to-noise considerations and statistical confidence in claiming presence rather than absence. LOD answers the binary question: "Is it there?" .
• Limit of Quantification (LOQ): The lowest concentration at which the analyte can not only be detected but quantified with predefined criteria for precision and accuracy. LOQ is where measurement uncertainty is acceptably small for the intended purpose. Some labs call this a reporting limit; methods or regulations may define the numerical criteria for LOQ differently (for example, a specified relative standard deviation or percent recovery) .
• Signal-to-noise ratio: Often used to estimate LOD and LOQ; historically, an S/N of ~3:1 has been used for LOD and ~10:1 for LOQ, but modern approaches favor statistically derived estimates that account for blank variability and method repeatability .
• Blank variability and method noise: If blank measurements vary, the threshold to claim detection must account for that variability. LOD is therefore a function of blank statistics and the distribution of low-level measurements .
• Precision vs. accuracy: Precision (repeatability) and accuracy (systematic bias from truth) both matter for LOQ. A number can be precise but biased, or unbiased but imprecise; LOQ normally requires both acceptable precision and acceptable bias in the relevant concentration interval .
• Calibration and linear range: Quantification depends on a calibration model that relates signal to concentration. LOQ should sit within the calibration range where the model is validated and residuals are acceptable. Outside that validated range, reported concentrations are extrapolations with unquantified risk .
• High relative uncertainty: Near the LOD, random noise and instrumental variability cause large relative uncertainty (percent error) even if the absolute signal is nonzero. Such uncertainty can render a numeric result uninformative for decisions that require a threshold or trend .
• Matrix effects: Sample matrices can suppress or enhance signals relative to calibration standards. If the method has not been demonstrated to control matrix effects at low concentrations, a detected signal may not translate reliably into a concentration without additional controls or corrections .
• Poor calibration at low end: Calibration curves are often less reliable near the low end because of heteroscedasticity (variance changing with concentration) and fewer low-level standards. If the calibration model is not validated down to the level of the detected signal, the concentration estimate is uncertain .
• Incomplete method performance evidence: Detection can be observed in one sample or run, but demonstrating LOQ requires replicate measurements, recovery studies, and precision/accuracy assessments across matrices. Without that evidence, a single detection is not sufficient to assert a quantified value with known uncertainty .
• Ask what limit was used: Was a reported number above the method LOQ, or is the lab reporting values between LOD and LOQ with caveats? Many laboratories will annotate low results as "<LOQ but detected" or similar; that distinction has practical implications for how the result should be interpreted .
• Inspect uncertainty and qualifiers: Reliable reporting should include method LOQ, LOD, and measurement uncertainty or qualifiers that explain confidence. If those are absent, the numeric value alone is incomplete evidence .
• Consider intended use: Different decisions tolerate different levels of uncertainty. Regulatory, clinical, or safety decisions often require concentrations above a validated LOQ, whereas screening or presence/absence surveillance may accept detection near LOD but should still note the limitations .
• Evaluate matrix and method documentation: Look for validation data showing LOQ performance in the same or closely related matrices. Methods validated in one matrix may not carry the same LOQ when applied to another without revalidation .