WoluTools
Find a tool

Microplate Assay Analyzer

ELISA 4PL and 5PL standard curve fitting online

Upload a plate-reader export as CSV or Excel. You get the fitted curve with its A, B, C and D parameters, back-calculated sample concentrations, and flags for high replicate CV, low R² and samples outside the standard range.

START HEREChoose data filesCSV, XLSX
Choose data files Download example plate (CSV)
No export file? Paste your values into the example sheet
  1. Download the example plate and open it in any spreadsheet program.
  2. Paste your readings over its rows, one row per well: well, sample, role, concentration, response. Role is standard, blank or sample; fill in the concentration only for standards.
  3. Save as CSV or XLSX and choose it with “Choose data files”.
Models
4PL · 5PL · linear · log-linear
Plate
96 or 384 wells · 1 file
File handling
Draft 2 hours · encrypted result 24 hours
FREE3 jobs a dayFiles up to 10 MB
PROUp to 200 jobs a day€12.99/month or €89.99/year
See plans

Larger files, up to the 64 MB tool limit, need Pro. Failed jobs are never counted.

  • No account needed
  • 3 free jobs a day
  • Research use only
EXAMPLE RESULT · FICTIONAL PLATE4PL · R² 0.9995
Fitted 4PL standard curve for the prepared example plateSigmoid curve of blank-corrected optical density against concentration on a log axis. Eight standards from 0.39 to 50 lie on the curve. Sample S14 falls below the lowest standard and S20 above the highest standard.00.511.522.530.1110100standard rangeConcentration (log scale)OD, blank-correctedC = 5.26S14S20S07
  • 8 standards, duplicates
  • Samples
  • Flagged
A response at zero
0.102
B slope
1.342
C EC50
5.26
D top plateau
2.811
Fitted by this tool from the example plate: 8 standards from 0.39 to 50, 3 blanks and 28 samples in duplicate.

Example result

28 samples read back through the fitted curve

Blank wells are averaged (0.061 OD here) and subtracted from every standard and sample. Each sample’s mean response is then converted to a concentration with the 4PL curve. In this run the fit converged after 249 iterations.

Interpolated unknowns 6 of 28 rows

SampleOD correctedCVConcentration
S040.4350.7%1.22
S121.8391.0%8.10
S032.1390.1%12.02
S071.43124.8%5.11
S140.1166.1%0.105below range
S202.7120.5%59.9above range

Out-of-range values stay in the table and are marked, so no sample disappears.

QC flags in this run 3 warnings

  • CV
    Replicate CV above your limitS07 · 24.8% against the confirmed 20%
  • Range
    Below the lowest standardS14 · 0.105 against 0.39
  • Range
    Above the highest standardS20 · 59.9 against 50

Also checked on every plate

  • R² below your minimum (default 0.99)
  • High CV on standard replicates
  • Standard means that do not rise or fall steadily
  • Responses the curve cannot convert, for example above the top plateau
  • A fit that did not converge
  • No blank wells (blank correction is then zero)

What you download

  • assay-results.csvPer sample: replicates, mean corrected response, CV, estimated concentration, in range yes or no
  • standard-curve.csvPer standard: mean response, CV, fitted response, residual
  • fitted-curve.csv201 points along the fitted curve for your own charts
  • PDF, XLSX, plotReport, evidence workbook with all parameters, curve image, plus JSON and a manifest

File layout

One row per well, five columns

Your column names do not have to match. In the workspace you type the names your export uses; the well and response columns are also found by common headers such as well, od or signal.

Example rows

wellsampleroleconcentrationresponse
A1STD1standard0.390.250
A3Blankblank0.061
D3S01sample1.194

The concentration, sample and role columns must be mapped before a job starts. Units follow your standards.

Settings you confirm

Curve model
4PL (at least 4 standard levels), 5PL (at least 5), linear or log-linear
Plate format
96 or 384 wells, up to 100,000 data rows
Maximum replicate CV
Default 20%
Minimum R²
Default 0.99
  1. 1
    Choose the fileOne CSV or XLSX export, or the example plate.
  2. 2
    Map and confirmColumns, model and limits stay visible before anything runs.
  3. 3
    Download the resultsTables, plot, PDF and workbook. Failed jobs are never counted.

Clear answers

ELISA curve fitting FAQ

Which plate-reader exports work here?

One CSV or XLSX table: up to 10 MB on the free tier; larger files, up to the 64 MB tool limit, need Pro. You map the well, response, concentration, sample and role columns yourself, so exports from different readers work without a fixed template.

Which curve models can I fit?

4PL, 5PL, linear and log-linear. You pick the model before the job starts. 4PL needs at least 4 distinct standard concentrations, 5PL at least 5. The fit uses the unweighted, blank-corrected mean of each standard level.

What do A, B, C and D mean?

A is the fitted response at zero concentration, D the response at the top plateau, C the EC50 and B the slope around it. 5PL adds an asymmetry parameter. All values and R² are written to the evidence XLSX and the JSON file.

What triggers a QC flag?

Replicate CV above the limit you confirm (standards and samples), R² below your minimum, standard means that do not change in one direction, samples outside the standard range, responses the curve cannot convert, and a fit that did not converge. Each flag names the sample or standard it came from.

Are out-of-range samples still reported?

Yes, but marked as such. A sample interpolated outside the standard curve appears as a review item instead of a concentration you can rely on.

Do I need an account?

No. The free tier gives 3 jobs a day without an account. Pro allows up to 200 jobs a day and larger files.

Can the result support a clinical decision?

No. The output is for research use. It records every selected column, model and parameter, and it offers no diagnostic or clinical interpretation.

Available now

Fit your own standard curve.

3 free jobs a day, no account needed. Files up to 10 MB on the free tier; drafts are kept 2 hours, encrypted results 24 hours.

Choose data filesSee the example result