Where many people struggle is not in understanding the principle but in interpreting the data that comes out. What does an OD450 reading of 1.27 actually mean? Is this sample positive or negative? How do you calculate the cutoff correctly? What does a standard curve tell you about your assay? When should you trust your result and when should you rerun? These are the practical questions this page addresses.
The Four ELISA Formats
Direct ELISA: The antigen is coated directly onto the microplate well. A primary antibody conjugated to an enzyme (most commonly horseradish peroxidase, HRP) is added. It binds to the antigen and the enzyme converts the substrate to produce colour. Simple, fast, but limited: the primary antibody must be directly labelled, and there is less signal amplification compared to indirect methods.
Indirect ELISA: The antigen is coated onto the well. An unconjugated primary antibody is added and binds to the antigen. Then a secondary antibody conjugated to HRP is added: it binds to the primary antibody. Because multiple secondary antibody molecules can bind each primary antibody, the signal is amplified compared to direct ELISA. Indirect ELISA is the format used to detect antibodies in patient samples: the patient's serum is the source of the primary antibody, and a commercial secondary antibody against human immunoglobulins is the detection reagent. Serology tests for HIV, SARS-CoV-2 antibodies, and Toxoplasma all use this principle.
Sandwich ELISA: A capture antibody is coated onto the well. The sample is added and the antigen of interest is captured by binding to the coated antibody. A second detection antibody (either directly HRP-conjugated or detected by a third antibody in double-sandwich format) is added to detect the captured antigen. Sandwich ELISA is highly specific and sensitive for antigen quantification because the antigen must bind two different antibodies simultaneously. It is used for cytokine detection, pathogen antigen quantification (for example, hepatitis B surface antigen), and many research applications.
Competitive ELISA: The sample antigen competes with a known quantity of labelled antigen for a fixed number of antibody binding sites. The higher the concentration of antigen in the sample, the less labelled antigen binds, and the lower the signal. Unlike the other formats, a higher OD in competitive ELISA means less antigen in the sample. This inverse relationship confuses many first-time users.
The Colour Reaction: What OD Values Mean
The most common enzyme-substrate system in clinical ELISA uses HRP and the substrate TMB (3,3',5,5'-tetramethylbenzidine). HRP converts TMB from colourless to blue. When the reaction is stopped with sulphuric acid, the colour changes from blue to yellow. The plate reader measures absorbance at 450 nanometres, which is the peak absorbance wavelength of the yellow TMB product. The concentration of yellow product is directly proportional to the amount of enzyme-bound antibody, which is proportional to the amount of target molecule in the sample.
An OD450 value of 0 means no colour, meaning no target was detected in that well. An OD450 of 0.05 to 0.15 is typical for a blank well (containing no enzyme at all). Negative controls typically give OD values in the range of 0.05 to 0.2. Positive controls or strongly positive samples produce OD values of 1.0 to 3.0 in most well-optimised ELISA formats. The plate reader can typically read accurately up to OD450 of about 3.0 to 3.5, and values above this may be saturated and inaccurate.
Calculating the Cutoff
The cutoff is the OD450 value above which a sample is classified as positive. Getting the cutoff right is the most important single calculation in interpreting an ELISA result.
The most common approach uses the mean OD of the negative controls plus a multiple of the standard deviation: Cutoff = Mean OD (negative controls) + N x SD (negative controls), where N is typically 2 or 3 depending on the desired stringency (N = 3 gives a more conservative, specific cutoff; N = 2 gives a more sensitive cutoff).
Alternatively, some commercial ELISA kits use a fixed multiplication factor: Cutoff = Mean OD (negative controls) x 2.1 (or another kit-specific factor). Always follow the kit insert's specified cutoff calculation rather than applying a generic formula to a commercial kit.
The sample-to-cutoff (S/CO) ratio is calculated for each sample: S/CO = Sample OD / Cutoff OD. An S/CO above 1.0 indicates a positive result. An S/CO below 1.0 indicates a negative result. Many laboratories define a borderline or equivocal zone (typically S/CO 0.9 to 1.1) within which samples are repeated or referred for confirmatory testing.
Standard Curve ELISA: Quantifying What You Find
In a sandwich ELISA used for quantification (for example, measuring the concentration of IL-6 in a cell culture supernatant), a series of wells containing known concentrations of the target analyte (the standard curve) is run alongside the unknown samples. The OD450 of each standard is measured and plotted against its concentration.
The resulting curve is not linear at high concentrations because the ELISA response saturates as antibody binding sites become occupied. The standard curve is fitted with a four-parameter logistic (4PL) model, which captures the sigmoidal shape of the response from low to high concentrations. The 4PL equation is: y = d + (a-d) / (1 + (x/c)^b), where a is the maximum asymptote (the highest OD achievable), d is the minimum asymptote (background), b is the Hill slope (steepness), and c is the EC50 (the analyte concentration at the midpoint of the curve). Fitting software (such as Excel with Solver, GraphPad Prism, or dedicated ELISA analysis software) fits this equation to the standard curve data.
Unknown concentrations are interpolated from the fitted curve by finding the x-value (concentration) that corresponds to their measured y-value (OD450). This interpolation only gives accurate results within the linear range of the curve, between the upper and lower limits of quantification. Samples with OD values outside this range need to be diluted (if too high) or re-extracted and concentrated (if too low) before accurate quantification is possible.
Validating Your ELISA Data
Before interpreting results, confirm that the run is valid by checking these quality indicators:
Positive control OD: Should fall within the expected range specified in the kit insert or your own historical data. An OD below the lower control limit means the assay underperformed: possible causes include degraded conjugate, incorrect incubation time, expired substrate, or a plate reader problem.
Negative control OD: Should be consistently low (typically below 0.2 for most clinical ELISAs). High negative control OD suggests non-specific binding, contamination of the negative control with antigen, or an excess of detection antibody. If negative control OD is too high, the calculated cutoff will also be elevated, potentially causing false-negative classifications.
Coefficient of variation (CV%) between duplicate wells: Duplicate sample wells should agree within CV% below 10 to 15 per cent for a well-optimised ELISA. Higher CV% suggests inconsistent pipetting, air bubbles, or uneven temperature distribution across the plate. Identify and re-run wells with high CV%.
Standard curve R²: Should exceed 0.99 for a well-fitted 4PL curve. An R² below 0.99 suggests the curve did not behave predictably, which may mean the standards were prepared incorrectly, degraded, or pipetted inaccurately.
Frequently Asked Questions
What does OD450 mean in ELISA?
OD450 is the optical density (absorbance) measured at 450 nanometres wavelength by the plate reader. After the HRP-TMB substrate reaction is stopped with sulphuric acid, the yellow product absorbs light maximally at 450 nm. The OD450 is proportional to the concentration of bound enzyme, which reflects the amount of target antigen or antibody present in the sample.
How do you calculate the cutoff for an ELISA?
The most common method is: Mean OD of negative controls plus 2 or 3 standard deviations of the negative controls. Some commercial kit inserts specify a multiplication factor applied to the mean negative OD. Always use the cutoff method specified for the specific ELISA format and kit you are using.
What is the S/CO ratio?
The sample-to-cutoff (S/CO) ratio is calculated by dividing the sample OD by the calculated cutoff OD. An S/CO above 1.0 indicates a positive result. Below 1.0 indicates negative. Values between 0.9 and 1.1 are often considered borderline and warrant repeat testing.
What is a 4PL curve?
A four-parameter logistic (4PL) curve is the mathematical model used to fit the sigmoidal shape of a quantitative ELISA standard curve. The four parameters describe the upper asymptote (maximum signal), lower asymptote (background), Hill slope (steepness of the linear region), and EC50 (midpoint concentration). Unknown sample concentrations are interpolated from the fitted curve.
What is the difference between direct and indirect ELISA?
In direct ELISA, the detecting antibody is directly conjugated to the enzyme. In indirect ELISA, an unlabelled primary antibody binds to the antigen, and a secondary antibody conjugated to the enzyme detects the primary antibody. Indirect ELISA amplifies the signal because multiple secondary antibody molecules can bind each primary antibody.
What does a competitive ELISA measure?
In competitive ELISA, sample antigen and labelled antigen compete for binding to a fixed amount of antibody. Higher sample antigen concentration means less labelled antigen binds, producing a lower signal. The relationship between signal and concentration is inverse: lower OD means more antigen in the sample.
Why is sulphuric acid used to stop the ELISA reaction?
Sulphuric acid stops the HRP-TMB reaction by denaturing the HRP enzyme and shifting the TMB colour from blue (the active product of HRP-TMB reaction) to yellow (the acid-converted form). The yellow colour is read at 450 nm. Stopping the reaction ensures all wells have the same reaction time before reading, giving consistent and comparable results.
What does high negative control OD mean?
High OD in negative control wells suggests non-specific binding: the detection antibody or conjugate is binding to the well surface or to components of the negative control sample rather than to specific antigen. Causes include insufficient blocking of the plate, too high a concentration of detection antibody, insufficient washing steps allowing unbound conjugate to remain, or genuine contamination of the negative control with the target antigen.
What is CV% and what is acceptable?
CV% (coefficient of variation percentage) measures the variability between duplicate or triplicate wells: CV% = (standard deviation / mean) x 100. For ELISA, CV% below 10 to 15 per cent between duplicate wells is generally acceptable. Higher CV% indicates inconsistent pipetting, air bubbles interfering with the reaction in specific wells, or temperature gradients across the plate.
How do you know if your ELISA standard curve is good?
A good 4PL standard curve fit has an R² above 0.99, the low-concentration standards are clearly above background (signal-to-noise ratio above 2), the highest-concentration standard has not reached saturation (the OD can still discriminate it from the second-highest), and the points follow the smooth sigmoidal curve shape predicted by the 4PL model without obvious outliers.