Most ELISA problems come from a handful of causes: poor washing, wrong incubation time or temperature, cold or badly prepared reagents, pipetting error and late or uneven reading. High background usually points to insufficient washing or too much detection reagent, while weak or no signal usually means a reagent was omitted, degraded or inhibited.
Troubleshooting matters because an ELISA result often drives a decision: releasing a batch labelled allergen-free, reporting a patient result or accepting a supplier. A run with a failed curve or noisy duplicates costs time, and a run that looks fine but is quietly biased costs far more.
ELISA formats in brief
Knowing the format tells you what normal looks like:
- Direct: antigen on the plate, detected by an enzyme-labelled primary antibody. Quick, with fewer steps.
- Indirect: an unlabelled primary antibody binds the antigen, then a labelled secondary antibody adds amplification. Common for antibody detection.
- Sandwich: a capture antibody binds the analyte and a labelled detection antibody binds a second site. Signal rises with concentration. Most protein and many allergen kits use this format.
- Competitive: sample analyte competes with labelled analyte for limited antibody, so signal falls as concentration rises. Used for small molecules and some processed-protein methods.
Most kits use horseradish peroxidase (HRP) with TMB substrate. The blue product turns yellow when an acid stop solution is added, and the plate is read at 450 nm, often with a reference wavelength to correct for plate imperfections.
ELISA troubleshooting table
| Problem | Likely causes | Fixes |
|---|---|---|
| High background | Insufficient washing or high residual volume; too much conjugate; poor blocking; contaminated or light-exposed substrate; substrate time too long; late reading | Check washer nozzles and residual volume; prepare wash buffer correctly; follow kit dilutions; use fresh, colourless TMB in the dark; time substrate exactly; read promptly |
| Weak or no signal | Reagent omitted or added in wrong order; expired or badly stored reagents; sodium azide in buffers (inhibits HRP); cold reagents; wrong wavelength | Use a step checklist; check expiry and storage; avoid azide with HRP; equilibrate to room temperature; confirm reader filter |
| High CV or poor duplicates | Pipetting error; bubbles; uneven washing; poor mixing; wells drying out | Calibrated pipettes, consistent technique, pre-wet tips; remove bubbles; service the washer; never let wells dry |
| Poor standard curve | Wrong reconstitution; degraded standard; serial dilution errors; wrong curve fit | Reconstitute exactly as directed; fresh dilutions; use 4PL or 5PL fit; check back-calculated recovery |
| Edge effects | Temperature gradients or evaporation at outer wells | Equilibrate plate and reagents; seal plates; do not stack plates in the incubator |
| Drift across plate | Slow reagent addition, so first and last wells incubate for different times | Multichannel pipette; add substrate and stop in the same order and pace |
| Hook effect (sandwich) | Very high analyte saturates capture and detection antibodies, giving falsely low signal | Test further dilutions; use a two-step protocol where the method allows |
The basics that fix most problems
Washing
Washing removes unbound material. Too little leaves conjugate behind and raises background; too aggressive can strip bound antibody. Use the number of cycles and soak times in the kit insert, check that every washer nozzle dispenses and aspirates, and tap the plate dry on clean absorbent paper after the final wash. Partly blocked nozzles are a frequent cause of patchy background.
Incubation time and temperature
Antibody binding depends on time and temperature. Use a timer for every step and the temperature stated in the method. Incubating above the specified range, for example near a warm instrument, can increase non-specific binding.
Reagent equilibration and preparation
Bring reagents and plates to room temperature before use, mix gently and prepare dilutions fresh. Cold substrate slows colour development, and wash buffer concentrate prepared at the wrong dilution is easy to miss.
Plate sealing and pipetting
Seal plates during incubations to prevent evaporation and contamination. Use calibrated pipettes, fresh tips for each standard and sample, and a consistent angle and speed. Add reagents to the side of the well without touching the coated surface.
Substrate, stop solution and reading
TMB is light-sensitive, so keep it in the dark and use only colourless substrate. Add stop solution in the same order and at the same pace as the substrate, then read within the time stated in the kit insert. Confirm the reader is set to the correct wavelength.
Curve fitting
ELISA standard curves are sigmoidal, so a straight-line fit across the full range biases results at the ends. A four-parameter logistic (4PL) fit suits most symmetrical curves, and a five-parameter logistic (5PL) fit can handle asymmetry. Report samples only within the validated range and dilute those above it.
Worked example: high background in an allergen ELISA in Dubai
A food testing lab in Dubai runs a sandwich ELISA for peanut protein in bakery products sold as peanut-free. On Monday, the zero standard and blank wells read well above the kit's acceptance limit, the curve is compressed at the low end and the lab cannot report results near the limit of quantification.
The analyst rejects the run and raises a nonconformity. The investigation works through the likely causes in order:
- Substrate: the TMB in the bottle is colourless and in date, which rules it out.
- Kit controls: the low control reads high on the same plate, so the problem lies in the run, not the samples.
- Wash buffer: the batch was correctly made up, as recorded in the preparation log.
- Washer: after maintenance, two nozzles in one manifold row are partly blocked. Those wells retain more liquid, and the unwashed conjugate raises the signal.
- Cross-contact: no peanut products are handled in the extraction area, and the reagent blank is also high, which suggests the issue is instrumental.
The lab cleans and primes the washer, verifies dispense and residual volume, and repeats the run with the kit controls. Background returns to normal and the controls fall within range. Because samples near the decision limit could have been affected, the lab re-tests every sample from the rejected run instead of reporting any of it. The CAPA adds a daily washer prime and visual nozzle check before each ELISA run.
This is also a good moment to check trend data. If the lab charts its kit controls, as described in our guide to internal quality control in medical labs, it can spot creeping background before a run fails outright.
Common mistakes
- Using reagents straight from the fridge.
- Stretching or shortening incubation times to fit a busy schedule.
- Mixing reagents from different kit lots without validation.
- Fitting a linear curve to sigmoidal data.
- Reporting samples above the top standard without dilution.
- Reading the plate long after adding stop solution.
- Ignoring a high zero standard because the controls just pass.
Key takeaway: Fix the basics before blaming the kit. Most ELISA failures come from washing, temperature, timing, pipetting and reading, so standardise them, check controls on every plate and investigate patterns rather than single wells.
Frequently Asked Questions
What causes high background in ELISA?
High background is usually caused by insufficient washing, too much conjugate or detection antibody, poor blocking, contaminated or light-exposed substrate, or overlong substrate incubation. Reading the plate late after adding stop solution can also raise optical density. Start by checking the washer and wash buffer, then reagent dilutions and substrate handling.
Why is there no signal in my ELISA?
No signal usually means a step failed. Common causes are a reagent left out or added in the wrong order, expired or badly stored reagents, sodium azide inhibiting HRP, very cold reagents or a wrong reader wavelength. Check the kit controls first: if they also fail, the problem is the run, not the samples.
What is the hook effect in sandwich ELISA?
The hook effect occurs when very high analyte concentrations saturate both capture and detection antibodies, so fewer sandwiches form and the signal is falsely low. It mainly affects one-step sandwich assays. Testing samples at more than one dilution, or using a two-step protocol where the method allows, helps detect and avoid it.
Should I use 4PL or 5PL for an ELISA standard curve?
Both are suitable for sigmoidal ELISA curves. A 4PL fit works well when the curve is roughly symmetrical, while 5PL adds a parameter for asymmetry. Use the model specified by the kit or your validated method and check that back-calculated standard concentrations recover acceptably before reporting results.
Next steps
If you run or review immunoassays, our Immunoassays and ELISA: Principles, Validation and Troubleshooting course covers assay design, curve fitting, acceptance criteria and investigation in practical depth.
For related reading, see method validation vs verification explained and internal quality control in medical labs. If you are unsure where to start, speak to a counsellor.
Looking for structured certification?
Explore Innoryx courses or speak to a counsellor about the right program for you.