How to Calculate IC50 from MTT and CCK-8 Cell Viability Assay Data
13 min read••By Chemical Express Thailand

Flat dose-response curves, edge-effect drift, and DMSO inconsistency across dilutions are the most common reasons IC50 values come back unreliable or irreproducible. A small setup error at the plate-layout stage compounds into a curve that software cannot fit. This guide walks you through every step of IC50 calculation cell viability, from plate seeding to 4PL curve fitting, so you finish with a documented IC50 and 95% confidence intervals you can defend.
Why Compute IC50 from Cell Viability and When It Matters
IC50 from a cell viability assay tells you the concentration of a compound that reduces metabolic activity by 50% in a given cell line. Use this metric when screening cytotoxicity, ranking compounds across multiple cell lines, or establishing starting doses for mechanistic follow-up studies.
Both MTT and CCK-8 measure metabolic activity as a colorimetric signal. That signal correlates with viable cell number, but only within a defined linear time window. Exceed it and the signal saturates, flattening the curve and pushing your IC50 estimate toward higher concentrations than the true value.
Assay choice, incubation time, and normalization method all influence the final IC50. Keep these parameters constant across experiments. This guide standardizes IC50 calculation cell viability so results remain comparable across plates and days. For a broader overview of cell culture analysis workflows, including media selection and culture maintenance, see the linked resource.
MTT vs CCK-8: Which assay fits your workflow?
- Lower reagent cost per plate
- Familiar protocol in most labs
- Requires formazan solubilization step with DMSO
- More susceptible to color interference from test compounds
- Best for routine cytotoxicity screening with no coloured compounds
- No solubilization step; read directly after incubation
- Higher sensitivity at low cell densities
- Reduced interference from many test compounds
- Preferred for colored or redox-active compounds
- Simpler workflow reduces handling variation
What You Need: Equipment, Reagents, and Software
Equipment
You need a CO2 incubator, biosafety cabinet, and a 96-well plate reader equipped with 450 nm and 570 nm filters. A reference filter at 630 nm or 650 nm reduces background noise from plate imperfections. A multichannel pipette and reagent reservoir improve throughput and reduce pipetting variation across rows.
Reagents and Controls
Prepare complete growth medium, sterile PBS, and high-purity DMSO as your vehicle. For the viability readout, you will use either an MTT reagent or a CCK-8 working solution. Include a positive cytotoxic control such as staurosporine to confirm your assay window detects full inhibition.
Controls on every plate must include: blank wells containing medium plus reagent only (no cells), a vehicle control matched to the highest DMSO concentration used across your dilution series, and optionally a no-cell background control for compound interference checks. Sourcing your metabolic assay kits from validated suppliers reduces lot-to-lot variation that quietly shifts IC50 values between experiments.
Software
You need a spreadsheet with a labelled plate map template and a curve-fitting tool. GraphPad Prism handles 4PL fitting natively with built-in CI reporting. Alternatively, use the drc package in R or scipy.optimize.curve_fit in Python for scripted, reproducible analysis across large compound sets.
Prepare and Seed Healthy Cells for a Reliable Signal
Cell health at seeding directly determines signal quality. Cultures past log phase or with viability below 90% extend lag times and compress your readout window. Confirm viability above 90% by trypan blue exclusion or an automated cell counter before proceeding.
- Prepare a single-cell suspension and count using a calibrated cell counter to confirm density and viability.
- Dilute to the target density. For most adherent lines, seed 5,000 to 10,000 cells per well in 100 µL of complete medium. Adjust density for faster-growing lines so the assay endpoint falls within the linear signal range.
- Seed interior wells of the 96-well plate according to the plate map below. Leave perimeter wells and fill them with sterile PBS or medium to buffer against edge evaporation.
- Incubate overnight at 37°C, 5% CO2 to allow cell attachment and recovery before compound dosing.
Follow the plate layout below to position blanks, vehicle, positive control, and dose series before seeding. Locking in the map before the experiment prevents ad hoc decisions that introduce positional bias.
| Well Position | Contents | Purpose | Replicates |
|---|---|---|---|
| Perimeter (Row A, Row H, Col 1, Col 12) | PBS or medium only | Evaporation buffer | All perimeter wells |
| B2 to B4 | Medium + reagent, no cells | Blank (background subtraction) | 3 |
| B5 to B7 | Cells + vehicle only | 100% viability reference | 3 |
| B8 to B10 | Cells + positive control (e.g., staurosporine) | 0% viability reference | 3 |
| C2 to G11 | Cells + test compound (8-10 concentrations) | Dose-response series | 3 per dose |
Treat Cells and Run the MTT or CCK-8 Viability Readout
Compound Dosing
- Prepare a half-log or log10 serial dilution of your test compound in complete medium. Cover at least 8 concentrations that bracket the expected IC50, typically spanning 4 to 5 orders of magnitude.
- Confirm that DMSO in the final dosing solution matches the vehicle control concentration exactly. Top dose should set the DMSO ceiling for the entire plate.
- Remove 100 µL of medium from each seeded well and replace with 100 µL of the appropriate treatment solution, maintaining a 1:1 dilution of pre-dosed volume.
- Return the plate to the incubator for the defined treatment duration, typically 24, 48, or 72 hours depending on the cell line and compound class.
Running the CCK-8 Readout
- At the assay endpoint, add 10 µL of CCK-8 reagent directly to each 100 µL well without removing medium.
- Incubate for 1 to 3 hours at 37°C, protected from direct light. Check the signal window empirically for your cell line before committing the full experiment.
- Read absorbance at 450 nm. Use 650 nm as a reference wavelength if your reader supports dual-wavelength correction.
Running the MTT Readout
- Add MTT solution to a final concentration of 0.5 mg/mL in each well.
- Incubate for 2 to 4 hours at 37°C until purple formazan crystals are clearly visible under the microscope.
- Carefully aspirate the supernatant without disturbing the crystal layer.
- Add 100 to 150 µL of DMSO to each well. Shake on an orbital mixer for 5 minutes until crystals dissolve completely.
- Read absorbance at 570 nm with a 630 nm reference if available.
The table below summarizes key operating parameters for each assay format.
| Parameter | CCK-8 (WST-8) | MTT |
|---|---|---|
| Reagent volume added per well | 10 µL | Volume to achieve 0.5 mg/mL final |
| Typical incubation time | 1 to 3 hours | 2 to 4 hours |
| Solubilization step required | No | Yes (DMSO, 100-150 µL) |
| Primary read wavelength | 450 nm | 570 nm |
| Reference wavelength | 650 nm | 630 nm |
If your test compound is colored or known to reduce tetrazolium reagents directly, include dedicated interference-blank wells: compound plus reagent, no cells. Subtract these values during analysis to isolate the true cellular signal.
Data Processing for IC50 Calculation Cell Viability
Normalize Absorbance and Compute Percent Viability
Start by subtracting the mean blank absorbance from every well on the plate. Then average the technical replicates within each dose group. Compute percent viability using this formula:
% Viability = 100 × (Sample absorbance - Blank) / (Vehicle control absorbance - Blank)
If viability values exceed 100% at sub-toxic doses, keep the raw values for fitting. Do not clip them before regression. If the overshoot is consistent and large, consider constraining the Top parameter during curve fitting and document that decision clearly in your methods.
Fit a 4-Parameter Logistic Curve and Extract IC50
Log10-transform all concentration values. Then fit the four-parameter logistic (4PL) model:
Y = Bottom + (Top - Bottom) / (1 + 10^((LogIC50 - X) × HillSlope))
Use these initial parameter estimates as starting guesses: Top near 100, Bottom near 0 to 10, LogIC50 at the mid-point of your concentration range, and HillSlope between 0.8 and 1.5. Start with unweighted least-squares fitting. If variance increases at high signal values, test 1/Y² weighting and compare residual plots.
Fitting by platform:
- GraphPad Prism: Choose nonlinear regression, select the built-in 4PL (log inhibitor vs normalized response) equation. Constrain Top and Bottom only when clearly justified by biology. Prism reports IC50, 95% CI, and HillSlope directly from the fit output.
- Microsoft Excel: Set up parameter cells for Top, Bottom, LogIC50, and HillSlope. Use Solver to minimize the sum of squared residuals. Compute confidence intervals via bootstrap resampling or asymptotic standard error estimation.
- R: Use the
drm()function in thedrcpackage with the LL.4 model. Extract parameters withsummary()and plot withplot(). Residual and leverage diagnostic plots are available natively. - Python: Use
scipy.optimize.curve_fit()with the 4PL function defined explicitly. Return the covariance matrix and compute 95% CI from the diagonal. Save residual plots for QC documentation.
Report Statistics and QC Acceptance
Accept a curve fit only when all four criteria are met: Top and Bottom estimates are physiologically plausible, residuals show no systematic pattern, at least two data points fall above the IC50, and at least two fall below. A curve that does not bracket the IC50 produces an extrapolated estimate with wide CI and low reliability.
When reporting IC50 values for comparative claims, state n as the number of independent biological experiments, not the count of technical replicates. Target a minimum of three independent experiments. Report IC50 as the geometric mean with 95% CI across those replicates.
Tips, Controls, and Precautions That Protect Data Integrity
Quick tips for reliable IC50 data
- Verify the linear signal window for each new cell line before running a full experiment. Perform a time-course with untreated cells and stop the assay before the signal plateaus.
- Use a reference wavelength (630 nm for MTT, 650 nm for CCK-8) to subtract plate and meniscus artifacts from each well independently.
- Pre-test DMSO tolerance for each cell line. Run a DMSO dose-response from 0.01% to 1% before starting compound work.
- Randomize compound positions across rows if running multiple plates on the same day. This distributes reader-lamp warm-up drift and pipetting fatigue across conditions rather than concentrating them in one dose group.
- Inspect each well visually for precipitate, bubbles, or color anomalies before reading. Remove air bubbles by brief centrifugation or gentle tapping.
- Clean the bottom of the plate with a lint-free tissue before reading. Fingerprints and condensation increase well-to-well absorbance noise.
- Minimize reagent light exposure during handling. Cover plates with foil during incubation and transfer steps.
- Log reagent lot numbers, incubation start and stop times to the minute, and plate reader calibration date for every run. Small timing differences compound across biological replicates.
Store WST-8-based reagents at 2 to 8°C, protected from light, and avoid repeated freeze-thaw cycles. MTT powder should be stored desiccated at 2 to 8°C; prepare fresh stock solutions monthly. A rising reagent blank or weakening signal at your positive control is the first sign of reagent degradation, and it will compress your IC50 before you notice the curve shape change.
When IC50 values are borderline or when distinguishing cytostasis from cytotoxicity matters for the study, add a fluorescence-based live cell imaging step to validate the viability readout. Live Cell Imaging Reagents let you confirm whether reduced metabolic signal reflects cell death or growth arrest, which changes the biological interpretation and the appropriate follow-up dose.
Apply the same normalization pipeline, incubation timing, and fitting constraints for every IC50 calculation cell viability run to maintain data continuity across batches and operators.
Troubleshooting Common IC50 Curve and Assay Issues
| Problem | Likely Cause | Fix |
|---|---|---|
| High background after blank subtraction | Reagent over-incubation; colored compound in medium; plate contamination | Shorten incubation to validated linear window; add compound-only blank wells and subtract; clean plate bottom before reading |
| Weak signal across all doses including vehicle control | Too few cells seeded; cells in poor health at seeding; reagent degraded | Increase seeding density; confirm viability above 90% before seeding; replace reagent lot and run fresh positive control |
| Edge effects with drifting Top plateau | Evaporation in perimeter wells affecting interior wells; uneven incubator airflow | Fill perimeter wells with PBS; seal plate with breathable membrane; rotate plate 180° at mid-incubation |
| Poor 4PL convergence or near-zero HillSlope | Insufficient concentration range; data do not bracket IC50; poor initial parameter estimates | Extend concentration range by 1 to 2 log units in both directions; re-seed initial guesses; check that at least 2 points fall above and below the IC50 |
| Color interference from test compound | Compound absorbs at assay wavelength or reduces tetrazolium directly | Add compound-plus-reagent-no-cell blank wells; subtract from each dose; switch to CCK-8 or ATP luminescence readout |
| Insoluble compound causing false low viability at high doses | Compound precipitates in aqueous medium above solubility limit | Inspect wells for precipitate before dosing; reduce top concentration; pre-dissolve in DMSO and verify solubility at final working concentrations |
Conclusion: From Raw Absorbance to a Defensible IC50
Disciplined plate setup, controlled incubation timing, and a consistent 4PL fitting workflow convert raw MTT or CCK-8 absorbance into a reproducible IC50 with confidence limits you can report. Every deviation from the standard pipeline, whether a variable DMSO load, a saturated signal, or an uncorrected compound interference, introduces systematic error that no statistical tool can fully remove after the fact.
Apply the normalization steps, fitting rules, and QC acceptance criteria in this guide consistently across experiments. That consistency is what makes IC50 values comparable between compounds, cell lines, and time points, which is the actual goal of IC50 calculation cell viability work.
Frequently Asked Questions
Recommended Products
All products below are available through Chemical Express Thailand. Submit an inquiry for pricing, availability, or technical guidance.
Recommended Products Available at Chemical Express
We deliver a wide range of IC50 products from Merck, MedChemExpress (MCE) and Elabscience and other international brands across Thailand. For any inquiry or purchase, fill out our Inquiry Form or chat with us instantly on Line.