A cell viability assay protocol is a step-by-step procedure for measuring how many cells in a sample are alive, dead, or metabolically active, and the choice of protocol matters enormously because different assays measure fundamentally different things. Some detect membrane integrity, others measure mitochondrial activity, and still others quantify total ATP. Each approach carries its own artifacts and blind spots, so a protocol that works flawlessly for one experiment can produce misleading results in another. Understanding what each assay actually measures, where it can go wrong, and how to set it up properly is what separates trustworthy data from numbers that look clean but mean little.
What Cell Viability Assays Actually Measure
The term “cell viability” sounds straightforward, but it covers several distinct biological states. A cell can have an intact membrane but be metabolically dormant. It can be metabolically active but committed to dying. It can be dividing rapidly or sitting in a quiescent phase. Different assay families tap into different readouts of “alive,” and that distinction is the first decision you make when designing a protocol.
Cell viability and proliferation assays are widely used in drug screening to determine whether test compounds affect cell growth or cause direct toxic effects.1PubMed. Cell Proliferation and Cytotoxicity Assays But the practical question researchers face is which readout matches their biological question. If you want to know whether a drug kills cells outright, membrane-integrity assays give the most direct answer. If you want to know whether surviving cells are still functioning normally, metabolic assays are more informative. And if you need continuous monitoring over hours or days, you need yet another category of technology.
Dye Exclusion Methods
Trypan blue exclusion is the oldest and simplest approach. You mix your cell suspension with the dye, load a small volume onto a hemocytometer or into an automated counter, and count. Live cells with intact membranes exclude the dye and appear bright under the microscope; dead cells take it up and turn blue. The whole process takes a few minutes and requires no specialized equipment beyond a microscope.
The simplicity is appealing, but the assay has well-documented limitations. Trypan blue can be incorporated by live cells after even a short exposure, so delays between adding the dye and counting introduce error.2PubMed Central. Trypan blue exclusion assay by flow cytometry Analyst-to-analyst variability is another consistent problem: two people counting the same sample on a hemocytometer routinely get different numbers, particularly when viability is in the ambiguous middle range where cells are faintly stained. Researchers working with both flat monolayer and three-dimensional cultures should be aware that the expected reliability of trypan blue counting can shift depending on the culture format, and measurement errors accumulate in ways that may or may not be acceptable for a given application.3PubMed Central. Cell Counting and Viability Assessment of 2D and 3D Cell Cultures: Expected Reliability of the Trypan Blue Assay
Despite these issues, trypan blue remains the go-to for routine passaging checks and quick viability estimates. The key protocol tips are to count within two to three minutes of adding dye, to standardize your counting squares across the lab, and to use automated counters when possible to minimize human bias.
Tetrazolium Salt Assays
MTT, MTS, XTT, and WST-1 are the workhorses of cell viability screening. They all rely on the same basic principle: metabolically active cells contain enzymes that convert a tetrazolium salt into a colored formazan product, and you read the color intensity on a plate reader. More color means more metabolic activity, which is used as a proxy for more viable cells.
MTT is the original version. Active mitochondria convert the yellow MTT substrate into a purple formazan product, and the color change can be quantified by spectrophotometry and correlated with cell number.4ScienceDirect. MTT Assay The formazan crystals formed in MTT assays are insoluble, so the protocol includes a solubilization step using DMSO or an acidified isopropanol solution before you can read the plate. That extra step is one reason newer variants like MTS and WST-1 became popular: they produce soluble formazan directly, letting you skip the dissolution and read the plate sooner. Some labs prefer WST-1 or XTT for kinetic experiments because you can take readings at multiple time points without killing the cells.
However, these newer variants are not always interchangeable with MTT. In a comparison study using colon cancer cells treated with sodium selenite, WST-1 and XTT failed to detect significant decreases in metabolic activity at concentrations where other assay types showed clear effects. In fact, lower concentrations of selenite appeared to increase metabolic activity as measured by WST-1 and XTT, with one cell line showing a roughly 40% increase above control values in the WST-1 assay.5Toxicology in Vitro. Antiproliferative effects of selenium compounds in colon cancer cells: Comparison of different cytotoxicity assays That kind of discrepancy is not a minor technical footnote. It means a compound that genuinely inhibits cell growth could look like it has no effect, or even a stimulatory effect, depending on which tetrazolium salt you chose.
Fluorometric and Luminescent Approaches
Resazurin-based assays (sold under names like alamarBlue and CellTiter-Blue) offer an alternative metabolic readout. Living cells reduce the blue, non-fluorescent resazurin molecule into pink, highly fluorescent resorufin.6PubMed. Investigation of the Alamar Blue (resazurin) fluorescent dye for the assessment of mammalian cell cytotoxicity You can measure the result either by fluorescence or by absorbance, giving you some flexibility with plate reader capabilities. The assay is also non-destructive in many protocols: after reading, you can wash the cells and proceed with another assay on the same well, which is useful for multiplexing.
ATP-based luminescent assays take a different approach entirely. Instead of measuring enzyme activity that produces color or fluorescence, they lyse the cells and quantify the total ATP released using a luciferase reaction. The light output correlates linearly with cell number across a wide range. Calibration curves for ATP bioluminescence assays show a linear correlation between light units and viable cell counts, with low detection limits.7PubMed. Application of BactTiter-Glo ATP bioluminescence assay for Mycobacterium tuberculosis detection The sensitivity is generally higher than tetrazolium assays, and the signal-to-noise ratio tends to be better at low cell densities. The trade-off is that ATP assays are destructive: they lyse everything in the well, so you get one reading per sample.
Compounds That Fool Your Assay
This is where a lot of published viability data goes quietly wrong. Tetrazolium and resazurin assays measure reduction chemistry, and any compound in your treatment wells that can reduce the reagent independently of cells will inflate your viability readings. The problem is especially severe with plant-derived extracts and polyphenols.
When antioxidant-rich extracts from Kakadu plum fruits were tested using both MTS-based and resazurin-based viability assays, higher extract concentrations appeared to significantly increase cell viability compared to lower concentrations. In cell-free control wells, the extracts directly reduced the assay reagent on their own, confirming that the elevated viability readings were artifacts of the compounds’ intrinsic reductive chemistry rather than genuine biological effects.8PubMed Central. Antioxidant-Rich Extracts of Terminalia ferdinandiana Interfere with Estimation of Cell Viability Any viability assay that depends on dehydrogenase activity is vulnerable to this kind of false positive when testing compounds with reducing potential.
Green tea polyphenols create a similar problem. In experiments with EGCG, the compound both protected mitochondria (increasing the ability of injured cells to reduce MTT) and directly converted MTT to formazan in cell-free conditions. The result was that MTT and MTS methods failed to detect antiproliferative effects at moderate concentrations, while ATP-based and DNA-based assays clearly showed growth inhibition.9PLoS ONE. Limitations of MTT and MTS-Based Assays for Measurement of Antiproliferative Activity of Green Tea Polyphenols Washing cells before adding the MTT reagent reduced the interference, which is a practical protocol modification worth adopting whenever you work with compounds that have known antioxidant activity.
The takeaway is that cell-free controls are not optional. Every time you run a viability assay with a new compound, you should include wells that contain the compound and the assay reagent but no cells. If you see color change or fluorescence in those wells, your assay is compromised, and you need to either switch to a different detection chemistry or build in a correction factor.
Nanoparticle Testing Deserves Its Own Protocol
Nanoparticles introduce a unique set of headaches for viability assays because they interact with optical readouts in ways that small-molecule drugs do not. Particles can absorb or scatter light at the same wavelengths used for detection, they can adsorb assay reagents onto their surfaces, and they can catalyze redox reactions that mimic or mask cellular activity.
Silver nanoparticles, for instance, directly interfere with MTS assay readouts. Depending on their coating and size, silver nanoparticles suppressed the measured absorbance of the formazan product by up to about 20%, independent of any actual cell killing. Citrate-coated particles caused more interference than PVP-coated particles of the same size, and smaller particles caused more interference than larger ones.10RSC Advances. Nanoparticles’ interference in the evaluation of in vitro toxicity of silver nanoparticles That means the assay could report that cells survived a treatment when the nanoparticles were actually the reason the signal dropped, or it could undercount the toxicity because the particles quenched the color from surviving cells.
A broader survey of nanoparticle-assay interactions tested silicon, cadmium selenide, titanium dioxide, and helical rosette nanotubes across six common assays. Every nanoparticle type interfered with at least one assay, producing either substantial overestimates or underestimates of toxicity.11PubMed Central. Widespread Nanoparticle-Assay Interference: Implications for Nanotoxicity Testing The interference was not limited to absorbance-based methods; fluorescence and even some luminescence readouts were affected, depending on the particle.
If you are working in nanotoxicology, the minimum protocol addition is to run at least two assays based on different detection principles and compare the results. If they agree, you have more confidence. If they disagree, you need to dig into which one is being fooled and by what mechanism.
Adapting Protocols for Three-Dimensional Cultures
Tumor spheroids and organoids are increasingly used because they better mimic how cells behave in tissue. But most viability assays were designed for flat monolayer cultures, and applying them to 3D structures without protocol adjustments leads to systematic errors.
The core issue is reagent penetration. In resazurin-based assays, tightly packed spheroids resist dye uptake, and the resulting signal underestimates the actual number of viable cells. Compact cell-cell interactions hamper resazurin penetration and its subsequent reduction to the fluorescent product, leading to lowered reduction activity relative to actual cell number.12PubMed. The Resazurin Reduction Assay Can Distinguish Cytotoxic from Cytostatic Compounds in Spheroid Screening Assays
ATP-based luminescent assays handle 3D cultures somewhat better because the lysis step breaks open the spheroid, releasing ATP from all cells including those in the interior. Even so, spheroids above a certain size create problems. Research on tumor spheroid models found that the linear relationship between spheroid size and luminescence signal broke down for spheroids larger than about 650 micrometers. The signal loss was attributed partly to incomplete reagent penetration and partly to the necrotic core that naturally forms in larger spheroids, reducing the number of viable cells at the center.13Scientific Reports. 3D tumor spheroid models for in vitro therapeutic screening: a systematic approach to enhance the biological relevance of data obtained
Practical adjustments for 3D protocols include extending incubation times to allow reagents more time to penetrate, using physical or enzymatic dissociation to break apart spheroids before adding viability reagents, and always including size-matched controls so you can distinguish drug effects from size-dependent signal loss.
Real-Time and Label-Free Methods
One limitation shared by nearly all the assays above is that they give you a snapshot at a single time point. You treat your cells, wait a set number of hours, add reagent, and read. If you want to know what happened between hour 4 and hour 48, you need separate plates for each time point, which multiplies your reagent costs and cell requirements.
Real-time cell analysis systems solve this by measuring electrical impedance across electrodes embedded in the bottom of specialized culture plates. As cells adhere, spread, and proliferate, they impede the flow of current, and the system records this continuously. The technique is label-free and allows real-time, continuous monitoring of cell adhesion, morphology, and rate of proliferation without adding any reagent.14Toxicology Reports. Real-time cell analysis system in cytotoxicity applications: Usefulness and comparison with tetrazolium salt assays This is especially valuable for capturing transient effects that a single endpoint would miss: a compound might cause cells to detach temporarily and then re-adhere, or it might slow growth for 12 hours before cells recover. Both of those dynamics are invisible to an endpoint assay.
A newer frontier is computational label-free imaging, where phase imaging and deep learning algorithms classify live and dead cells from standard microscopy images without any staining. One approach using quantitative phase imaging reported roughly 95% accuracy in identifying live versus dead cells across different culture types.15Nature Communications. Live-dead assay on unlabeled cells using phase imaging with computational specificity The method is still mostly in academic development, but it points toward a future where viability assessment could be built into routine imaging workflows without consuming any reagents.
Plate Layout and Edge Effects
Even when the assay chemistry is perfectly suited to your experiment, a sloppy plate layout can wreck your data. The most common trap is the edge effect: wells around the perimeter of a 96-well plate experience slightly more evaporation during incubation, which concentrates the medium, changes osmolarity, and shifts metabolic readouts. Research optimizing cancer drug sensitivity screens observed elevated resazurin-based absorbance values in perimeter wells, even in plates designed to minimize evaporation. The effect appeared consistently across untreated cells, vehicle-treated cells, and drug-treated cells.16Scientific Reports. Optimization of cell viability assays to improve replicability and reproducibility of cancer drug sensitivity screens
The standard workaround is to fill the outermost wells with medium only (no cells, no treatment) and confine your experimental conditions to the interior wells. You lose 36 of your 96 wells this way, but the trade-off in data quality is worth it. Other protocol elements that improve reproducibility include seeding a defined number of cells per well (rather than estimating), giving cells a full day to attach before adding treatments, and reading the plate at a consistent incubation time post-reagent addition. Small differences in these steps compound across replicates and between experimenters.
Regulatory Standards for Viability Testing
If your viability testing feeds into a regulatory submission, particularly for medical devices and biomaterials, you operate under ISO 10993-5, which governs in vitro cytotoxicity testing as part of the broader biological evaluation of medical devices. The current version of this standard specifies four metabolic assays that primarily rely on colorimetric methods and colony formation.17PubMed Central. Revisiting ISO 10993-5 In Vitro Cytotoxicity Standard Tests for Evaluation of Extracellular Matrix-Based Biomaterials The standard was published in 2009 and has not been substantially updated to reflect advances like real-time impedance monitoring or label-free imaging, so labs performing regulatory testing are typically locked into older assay formats whether or not a newer method would be more informative for their material.
For non-regulatory applications, your assay choice is wide open, but it is still good practice to report enough protocol detail that someone else can replicate your experiment. That means specifying the assay kit and catalog number, cell seeding density, treatment duration, reagent incubation time, plate reader settings, and how you handled your controls. Published viability data that omits these details is frustratingly common and essentially unreproducible.
Viability Assessment for Primary Cells and Microbes
Most protocol guides assume you are working with a hardy, immortalized cell line like HeLa or HEK293. Primary cells, especially neurons, are more fragile and respond differently to assay conditions. Protocols designed for primary neuronal cultures typically include multiple complementary approaches: membrane-integrity assays like trypan blue and propidium iodide exclusion, metabolic assays like MTT, and release assays that measure lactate dehydrogenase leaking from damaged cells into the culture medium.18PubMed. Assessment of cell viability in primary neuronal cultures Running more than one assay on primary cultures is not overkill; it is a practical response to the fact that these cells die more easily during handling, and any single readout could be confounded by the culture conditions.
Bacterial viability assessment works from some of the same principles but adds layers of complexity. Bacteria can enter a “viable but non-culturable” state where they are metabolically active but will not grow on a standard agar plate. Viability strategies for bacterial pathogens are generally organized around three criteria: whether the bacteria can be cultured, whether they are metabolically active, and whether their membranes are intact. Methods based on membrane integrity can detect viable bacteria that culture-based approaches miss entirely.19PubMed Central. Recent Methods for the Viability Assessment of Bacterial Pathogens: Advances, Challenges, and Future Perspectives If your work involves antimicrobial testing, relying on colony counts alone can severely underestimate the number of surviving organisms, because a fraction of the survivors may be in that dormant-but-alive state.
Choosing and Validating the Right Protocol
No single viability assay is best for all situations, and the most common mistake is picking an assay because it is familiar rather than because it fits the experiment. A useful decision framework starts with three questions: What biological state are you trying to measure (membrane integrity, metabolic activity, total biomass, proliferative capacity)? What compounds or materials are you testing, and could they chemically interfere with the assay reagents? And what throughput and time resolution do you need?
If you are screening thousands of compounds in a drug discovery pipeline, ATP luminescence in 384-well format gives you speed and sensitivity. Miniaturized assays in high-density plates accelerate early-stage screening, though accurate fluid handling at sub-microliter volumes remains a technical challenge.20PubMed. Challenges and solutions to ultra-high-throughput screening assay miniaturization: submicroliter fluid handling If you are evaluating a biomaterial for a device submission, ISO 10993-5 limits your menu. If you are testing plant extracts or nanoparticles, you need to run cell-free controls and ideally confirm results with a second assay based on different chemistry.
Validation does not need to be elaborate. At minimum, confirm that your assay signal scales linearly with cell number across the range you plan to use, run cell-free interference controls with your treatment compounds, include positive controls (a known toxic agent at a concentration that reliably kills most cells), and document your protocol thoroughly enough that a colleague unfamiliar with the assay could reproduce your experiment without asking you questions. These steps catch most of the errors that quietly undermine viability data, and they take far less time than repeating an experiment because the first round produced numbers nobody trusts.