Competitive vs Sandwich ELISA: Should You Subtract the Blank?

The problem

A researcher was running two ELISA kits in parallel — a Serotonin assay and a Zonulin assay — and applied the same data-analysis workflow to both. After subtracting the blank from the raw optical density (OD) readings, every calculated value on the Serotonin plate came out negative. The researcher suspected that blank subtraction might not apply here and reached out to confirm how the data should be handled.

The instinct was correct. The two kits use fundamentally different assay formats, and only one of them calls for subtracting the blank.

Snapshot of the data

Here is what a competitive standard curve looks like, and what happens when the blank is subtracted from it. As concentration rises, OD falls — so subtracting the blank (the highest OD) leaves every point below zero.

Values are illustrative and shown only to demonstrate the pattern.

The second table contrasts the two formats side by side — the same distinction that separates a competitive kit from a sandwich kit.

Values are illustrative and shown only to demonstrate the pattern.

Why it happens

In a sandwich ELISA, the target is captured between two antibodies, and more target means more signal — OD rises with concentration. The blank (no target) is the lowest reading, so subtracting it simply removes background. That is why sandwich protocols routinely say “subtract the blank.”

A competitive ELISA works the other way. The sample competes with a labelled antigen for a limited number of antibody sites, so more target displaces more label and produces less signal. The zero standard (B0) has nothing to compete and therefore binds the most label, giving the highest OD on the plate. Subtract that peak value from every other well and the entire curve collapses below zero — exactly what the researcher saw.

Our analysis

Reviewing the raw data, the sample ODs and the shape of the competitive curve looked reasonable — the negatives were purely an artefact of blank subtraction, not a failed assay. Once the blank subtraction was removed, the values fit the standard curve as expected.

One detail was worth a second look. On a healthy competitive curve, the top of the curve (B0, the zero standard) should read around an OD of 2.0. The researcher’s B0 sat noticeably lower. In discussing the prep, two points stood out: the reconstituted standard had been left to equilibrate for only a couple of minutes before dilution, and the blank had been prepared using the sample-dilution buffer. A B0 that reads low usually points back to standard reconstitution or to the choice of diluent for the zero/blank well.

Root cause

What we recommended

Key takeaway

Frequently asked questions

How do I know if my kit is competitive or sandwich?

The datasheet states the format explicitly. As a rule of thumb, small molecules and hormones (serotonin, cortisol, and similar) are usually competitive, while larger proteins and cytokines (zonulin, interleukins) are usually sandwich.

Why are my competitive ELISA values negative after subtracting the blank?

Because the blank is the highest OD in a competitive assay. Subtracting it makes every other well negative. Re-analyse without blank subtraction.

What OD should the blank (B0) give in a competitive assay?

Around 2.0 is typical for a well-prepared curve. A markedly lower B0 usually means the standard was not fully reconstituted or the wrong diluent was used for the zero well.

Do I still need a blank on a competitive plate?

Yes — run it as the zero standard (B0) to anchor the top of the curve. You just don’t subtract it from the other wells.

Serotonin ELISA Kit

SKU: UNFI0018 · Competitive ELISA

Detection range	1.563–100 ng/mL
Sensitivity	0.938 ng/mL
Sample types	Serum, plasma, cell culture supernatant, cell/tissue lysate
Price	€649

View product →

This article is based on an anonymized technical-support case and is provided for general educational purposes. Optical density values are illustrative. Always follow the protocol and specifications in your kit’s current datasheet, as ranges and reagents may vary by lot and product version.