An enzyme-linked immunosorbent assay, or ELISA, uses binding reagents, usually antibodies, to convert the presence of an analyte into an optical or other measurable signal. It can be sensitive and scalable. The result is only as specific as the antibodies, calibration model, sample matrix, and validation.
A number reported in picograms per milliliter is not direct molecular counting. It is an estimate inferred from how the sample behaved relative to standards.
Common formats create different evidence
In a sandwich ELISA, a capture antibody binds the analyte and a second detection antibody recognizes another epitope. This format can be highly specific because two binding events are required.
Competitive ELISAs measure how sample analyte competes with a labeled or immobilized reagent. Signal often decreases as analyte increases. They can work for small molecules or analytes with only one accessible epitope, but the inverse curve can be less intuitive.
Indirect formats measure antibodies against an antigen rather than antigen concentration. A positive anti-drug-antibody result is not the same as a measured drug level.
The format should be named before the result is interpreted.
The calibration curve is usually nonlinear
Many ELISAs use four- or five-parameter logistic models to relate signal to concentration. The curve has flatter regions near the low and high asymptotes and a steeper middle region.
Precision and sensitivity are not uniform across the curve. Samples should be diluted into the validated quantitation range rather than extrapolated far beyond the highest standard.
The report should show standard concentrations, replicate signals, curve model, weighting, acceptance criteria, and back-calculated standard performance.
Antibody specificity is empirical
An antibody raised against a peptide can bind related sequences, fragments, precursors, modified forms, or carrier-associated epitopes. Cross-reactivity depends on concentration and matrix, not only sequence similarity.
Specificity testing should challenge plausible related molecules and abundant matrix components. “No cross-reactivity” means no qualifying response under tested conditions and limits; it does not prove universal exclusivity.
For endogenous analytes, the assay may detect multiple isoforms. The report should define what molecular forms the result represents.
Matrix effects can suppress or enhance signal
Serum, plasma, cell-culture medium, formulation buffers, and tissue extracts contain proteins, lipids, salts, heterophilic antibodies, and other substances that can alter binding or detection.
Spike-recovery tests add known analyte to the matrix and measure how much is recovered. Dilution parallelism examines whether serially diluted samples behave in parallel with the calibration curve. Both help show that concentration estimates remain valid in real samples.
Good recovery in buffer does not establish suitability in plasma.
Hook effect can make a high sample look low
In some sandwich assays, an extremely high analyte concentration saturates capture and detection antibodies without forming the expected sandwich, producing a falsely low result. This high-dose hook effect can be detected by testing additional dilutions.
A report that automatically dilutes only samples with high initial signal can miss hooked samples. Validation should establish whether and where the effect occurs.
Values below the range are not zero
The lower limit of quantitation is the lowest concentration meeting predefined accuracy and precision. A detectable signal below that point may be reported as below quantitation rather than assigned an exact concentration.
Substituting zero or half the lower limit in statistical analysis changes the data and should be justified. When many results are censored, methods designed for censored values may be needed.
Likewise, a value above the upper limit should be diluted and reanalyzed if sample stability and dilution validity permit.
Pre-analytical handling can dominate
Analytes may degrade, adsorb to tubes, bind soluble receptors, or change through repeated freeze-thaw cycles. Collection tube, anticoagulant, clotting time, processing delay, storage temperature, and freeze-thaw history should be standardized.
Plate position, incubation time, washing, reagent temperature, and substrate timing create analytical variation. Controls across the plate help detect drift and edge effects.
A report checklist
Look for:
- assay format and antibody identities;
- standard material and assigned concentration;
- matrix and sample preparation;
- calibration range, model, and weighting;
- lower and upper quantitation limits;
- accuracy, precision, and selectivity;
- dilution parallelism and spike recovery;
- hook-effect assessment;
- plate controls and replicate rules;
- handling of below-range, above-range, and repeat results.
For regulated bioanalysis, stability and lot-to-lot reagent performance also belong in the validation package.
What ELISA cannot establish alone
An ELISA result does not independently prove intact molecular identity, biological activity, sequence, or purity. Immunoreactivity can persist in a fragment or modified molecule, while an active molecule can be underestimated if its epitope is masked.
Orthogonal LC-MS, functional assays, or another immunoassay with different epitopes can clarify difficult claims.
Bottom line
ELISA is a binding-based comparative measurement. Trust depends on the standard, antibodies, matrix validation, and curve range. The most defensible report shows how the sample fits the validated measurement system rather than presenting concentration as an instrument-independent fact.
Sources
- U.S. Food and Drug Administration. Bioanalytical Method Validation Guidance for Industry.
- International Council for Harmonisation. M10: Bioanalytical Method Validation and Study Sample Analysis.
- European Medicines Agency. ICH M10 Bioanalytical Method Validation.
- Findlay JWA, Dillard RF. Appropriate calibration curve fitting in ligand binding assays. AAPS Journal. 2007.
- U.S. Food and Drug Administration. Immunogenicity Testing of Therapeutic Protein Products, Developing and Validating Assays.
Continue reading
Editorial responsibility and corrections
Published by Serum Research. Sources checked October 1, 2026. This is educational material; no independent scientific, clinical, statistical, analytical, or regulatory review is claimed. Read how we review evidence or send a correction.
For Research Use Only. Not for Human Consumption. This page does not provide instructions for medical use or establish the identity, quality, safety, or equivalence of a catalog material.