CellTiter-Glo is a luminescent cell viability assay that measures the amount of adenosine triphosphate (ATP) in a sample as a direct proxy for the number of metabolically active cells. Developed by Promega, it relies on a single-reagent, add-and-read format that has made it one of the most widely adopted viability platforms in drug screening, cytotoxicity testing, and basic cell biology. The protocol is straightforward, but getting reliable, artifact-free data from it requires understanding its chemistry, its optimization demands, and the specific situations where it can mislead you.
How the Chemistry Works
The assay uses the firefly luciferase reaction. When you add CellTiter-Glo reagent to your cells, it lyses them and simultaneously provides the enzyme luciferase along with its substrate, luciferin. In the presence of ATP released from viable cells, luciferase catalyzes a reaction that produces oxyluciferin and, critically, light. The amount of luminescent signal is proportional to the amount of ATP present, which in turn is proportional to the number of viable cells in the well. Dead cells do not maintain ATP pools, so they contribute little to no signal.
This approach belongs to a broader family of viability methods that use beetle luciferase to measure ATP as a marker of living cells.1SpringerLink. Cytotoxicity testing: measuring viable cells, dead cells, and detecting mechanism of cell death The luminescent readout offers a practical advantage over colorimetric or fluorescent alternatives: luminescence has essentially no background from the cells or media themselves, which gives the assay a wide dynamic range and high sensitivity. That low background is a big part of why CellTiter-Glo became a default choice for high-throughput screening campaigns where thousands of compounds need to be evaluated quickly.
The Basic Protocol
Running CellTiter-Glo in a standard plate format is simple on the surface. You culture your cells in multiwell plates, apply your treatments for the desired duration, then add an equal volume of CellTiter-Glo reagent directly to the wells. The reagent lyses cells and initiates the luminescent reaction simultaneously. After a brief mixing step (typically orbital shaking for about two minutes to promote complete lysis), you let the plate sit at room temperature for around ten minutes to stabilize the signal, then read it on a luminescence-capable plate reader.
The glow-type signal the assay produces is one of its most user-friendly features. Unlike flash-type luminescent assays where the signal peaks and decays within seconds, CellTiter-Glo generates a stable signal with a half-life of more than five hours. This means you do not need an injector on your plate reader, and you have a comfortable window to read your plates without worrying about signal decay introducing variability between the first and last wells.
Why Optimization Matters More Than the Protocol Sheet Suggests
The reagent addition step is easy. The part that separates reliable data from misleading data is the optimization that happens before you ever add the reagent. The most important variable is seeding density. You need cells to be within the linear range of the assay at your experimental endpoint, meaning the relationship between cell number and luminescent signal is still proportional. If your cells have grown to confluency and are packed too tightly, the ATP signal plateaus and you lose the ability to distinguish between, say, 80% viability and 100% viability. Subtle drug effects disappear into that plateau.
Researchers working with prostate cancer cell lines demonstrated this rigorously by combining live-cell imaging with endpoint CellTiter-Glo readings. They found that cells needed to stay below roughly 70% confluence at the experimental endpoint for the luminescent signal to remain within the detectable linear range. For one cell line (BPH-1), this meant seeding just 500 cells per well in a 96-well plate with 2% serum-supplemented media for a 72-hour drug screening experiment. Each cell line in their panel required its own optimized seeding density and media formulation to meet the same criteria.2PLoS ONE. Automation, live-cell imaging, and endpoint cell viability for prostate cancer drug screens The takeaway is that you cannot simply pick a round number for seeding density and assume it will work across different cell types or treatment durations.
Other optimization considerations include ensuring your vehicle-treated control cells continue proliferating throughout the experiment (so you have a true baseline of normal growth), and accounting for edge effects in multiwell plates where evaporation can cause wells along the perimeter to behave differently from interior wells. Many experienced users leave the outermost wells empty or fill them with media only.
High-Throughput Screening Performance
CellTiter-Glo’s reputation was largely built in high-throughput screening (HTS) environments, where you need an assay that is both robust and automatable. Two validation studies in antiviral screening illustrate why. In a 384-well assay designed to identify potential influenza antivirals, the luminescent readout achieved Z-factor values above 0.8, signal-to-background ratios greater than 30, and signal-to-noise ratios above 10, all well above the thresholds considered acceptable for HTS.3PubMed. A cell-based luminescence assay is effective for high-throughput screening of potential influenza antivirals A separate screen for respiratory syncytial virus inhibitors reported similarly strong metrics, with Z-factors above 0.8 and signal-to-background exceeding 35.4PubMed Central. A cell based high-throughput screening approach for the discovery of new inhibitors of respiratory syncytial virus
To put those numbers in context, a Z-factor above 0.5 is generally considered a good assay for screening; above 0.8 signals an excellent one with clear separation between positive and negative controls. CellTiter-Glo consistently hits those marks, which is why it remains a workhorse in pharma and academic screening facilities. The assay has also performed well outside mammalian cell systems: in a screen against the parasitic flatworm Schistosoma mansoni, CellTiter-Glo identified 100% of the hit compounds that were found by the gold-standard method of microscopic evaluation, while producing signals at least three-fold above background noise.5PubMed Central. Fluorescence/luminescence-based markers for the assessment of Schistosoma mansoni schistosomula drug assays
How It Compares to Other Viability Assays
No single viability assay is perfect, and understanding where CellTiter-Glo sits relative to alternatives helps you decide when to trust it and when to reach for something else. A systematic comparison in glioma stem-like cells tested ATP-based luminescence (CellTiter-Glo) against NADH-based assays and real-time confluency monitoring. The ATP and NADH assays produced comparable results across treatments, but the ATP-based readout had smaller standard deviations and a more direct signal, making it easier to interpret.6PubMed Central. A Systematic Comparison Identifies an ATP-Based Viability Assay as Most Suitable Read-Out for Drug Screening in Glioma Stem-Like Cells The researchers concluded that combining ATP luminescence with confluency monitoring provided the most specific and reproducible readout for drug screening on primary glioma stem cells.
A separate comparison tested CellTiter-Glo, resazurin reduction (a metabolic assay), and nuclei enumeration (a direct cell-counting method). Both CellTiter-Glo and resazurin showed higher viability readings compared to stained nuclei counts.7SLAS Discovery. A Comparison of Real-Time and Endpoint Cell Viability Assays for Improved Synthetic Lethal Drug Validation This discrepancy matters. Because CellTiter-Glo measures metabolic activity (ATP content) rather than physically counting cells, it can overestimate viability when a drug slows proliferation without immediately killing cells. The remaining cells may be metabolically healthy and producing plenty of ATP, even though there are fewer of them than in the control. For this reason, pairing CellTiter-Glo with a direct cell counting method or real-time imaging is a common and well-justified practice in drug discovery.
The 3D Culture Problem
As cell biology has moved toward three-dimensional culture systems like spheroids, organoids, and hydrogel-embedded cultures, researchers have naturally tried to apply CellTiter-Glo to these models. Promega even developed a dedicated “CellTiter-Glo 3D” formulation with stronger lytic capacity designed to penetrate dense cell aggregates. But the results have been mixed.
A comparative study tested CellTiter-Glo, CellTiter-Glo 3D, an MTS colorimetric assay, and PrestoBlue across different hydrogel formulations. All of the assays showed inaccuracies when compared to direct microscopic assessments of cell viability in the 3D systems. The researchers concluded that microscopic imaging should be used alongside any chemical viability assay for validation when working in three-dimensional cultures.8Mary Ann Liebert, Inc., publishers. Cell Viability Assays in Three-Dimensional Hydrogels: A Comparative Study of Accuracy The core issue is that reagent penetration into thick 3D structures is inherently uneven. ATP from cells deep inside a spheroid may not be fully released, or the luminescent reagent may not reach them completely, leading to an undercount of viable cells. The hydrogel material itself can also interfere with signal transmission or sequester reagent components.
If you are working with 3D cultures, treat CellTiter-Glo results as a relative measure between treated and untreated conditions rather than an absolute quantification of cell number. And always confirm findings with imaging or an orthogonal assay.
Artifacts and Compound Interference
One of the less-discussed pitfalls of any luciferase-based assay is that test compounds can directly inhibit the luciferase enzyme, producing a false drop in signal that looks like cytotoxicity but is actually a chemical artifact. This is especially relevant in large screening campaigns where compound libraries include diverse chemical scaffolds, some of which will inevitably interact with the enzyme’s active site.
Extensive profiling of nearly 200,000 compounds against firefly luciferase from Photinus pyralis (the wild-type enzyme used in many older assays) found that about 0.9% of the library inhibited the enzyme at concentrations below 10 micromolar. CellTiter-Glo uses an engineered variant called Ultra-Glo luciferase, which was developed specifically to reduce this problem. Against Ultra-Glo, only about 0.1% of compounds showed inhibition in the same concentration range, and the most potent inhibitor was roughly twelve-fold weaker than the most potent inhibitor of the wild-type enzyme.9PubMed Central. A basis for reduced chemical library inhibition of firefly luciferase obtained from directed evolution Ultra-Glo achieves this by eliminating binding interactions outside the luciferin pocket that the wild-type enzyme is susceptible to.
Even with this improvement, compound interference has not been eliminated entirely. Best practice in screening is to run counter-screens: test your hit compounds directly against the luciferase enzyme in a cell-free system. If a compound inhibits the enzyme itself, it is an assay artifact, not a biologically meaningful hit. Promega and other vendors sell cell-free luciferase detection kits specifically for this purpose.
When ATP Does Not Mean What You Think It Means
The foundational assumption of CellTiter-Glo is that ATP content reflects cell viability. In most standard culture conditions, this holds well. But there are situations where ATP levels and viability diverge, and the assay will give you a misleading answer.
Drugs that directly target metabolic pathways are the most obvious concern. A compound that inhibits mitochondrial respiration or glycolysis can crash ATP levels rapidly while cells are still structurally intact and potentially recoverable. The glioma stem cell comparison study flagged this explicitly, noting that drugs interfering with cell metabolism require alternative measurement techniques like confluency monitoring to get an accurate picture.6PubMed Central. A Systematic Comparison Identifies an ATP-Based Viability Assay as Most Suitable Read-Out for Drug Screening in Glioma Stem-Like Cells
The reverse scenario also exists. A study on cultured rat astrocytes found that cells maintained high ATP concentrations for at least six hours even in the complete absence of glucose and amino acids. After 24 hours of nutrient deprivation, ATP had only declined to about 30% of its initial value, and the cells were still viable.10PubMed Central. Endogenous Energy Stores Maintain a High ATP Concentration for Hours in Glucose-Depleted Cultured Primary Rat Astrocytes Cells can buffer ATP through stored glycogen, fatty acid oxidation, and other reserve pathways. This means a treated population might look healthier than it actually is if you read the plate before ATP reserves have been fully depleted.
One creative approach to exploit this gap intentionally involves pre-treating cells with the glycolysis inhibitor 2-deoxyglucose (2-DG) before adding test compounds and reading with CellTiter-Glo. By forcing cells to depend on mitochondrial metabolism, researchers made the assay dramatically more sensitive to compounds targeting mitochondrial function. In one implementation, the number of active compounds identified jumped from 3 out of 30 to 13 out of 30 when 2-DG was included, and compound potency increased roughly tenfold.11PubMed Central. Development of a Rapid In Vitro Screening Assay Using Metabolic Inhibitors to Detect Highly Selective Anticancer Agents The assay maintained a Z-factor of 0.805, confirming that the metabolic manipulation did not compromise screening quality.
Multiplexing With Other Assays
Because CellTiter-Glo is a lytic, endpoint assay, it destroys the sample. You cannot go back and run another assay on the same cells afterward. This has driven interest in multiplexing strategies where you pair CellTiter-Glo with non-lytic or real-time assays that can be read first, before adding the CellTiter-Glo reagent as a terminal step.
A common approach involves reading a fluorescent cytotoxicity marker (like CellTox Green, which labels dead cells by staining compromised membranes) first, then adding CellTiter-Glo to the same wells to get a viability measurement. Testing this combination showed that the CellTox Green dye caused a roughly 20-23% proportional reduction in CellTiter-Glo signal. Importantly, this reduction was consistent across all compound concentrations and did not shift the calculated IC50 values for the tested compounds.12Promega Corporation. Gain More Informative Data by Multiplexing a Fluorescent Real-Time Cytotoxicity Assay with Luminescent, Fluorescent or Colorimetric Viability Assays So the absolute signal drops, but the relative differences between conditions remain intact.
More elaborate multiplexing schemes measure several endpoints from parallel plates. One framework measured cell viability, caspase-3/7 activity (a marker of apoptosis), lactate dehydrogenase release (a marker of membrane integrity), ATP content, and two additional enzyme markers from paired multiwell plates.13PubMed. Multiplexing cell viability assays This kind of multi-endpoint approach helps distinguish whether a compound kills cells by triggering programmed cell death, by physically rupturing membranes, or by some other mechanism. CellTiter-Glo alone tells you cells are dying; the additional markers help explain how.
Plate Reader Selection and Crosstalk
A factor that receives surprisingly little attention in protocol guides is how much your plate reader itself influences your data. Luminescence readings are highly dependent on detector sensitivity, and different instruments can produce dramatically different raw signal values from the same plate. Beyond absolute sensitivity, crosstalk between adjacent wells can be a problem, especially in 384-well and 1536-well formats where wells are physically close together. Crosstalk occurs when light from a bright well bleeds into the detector’s reading of a neighboring dim well, inflating the apparent signal in the dim well.
Promega’s own testing across different microplate readers showed measurable differences in crosstalk performance, which they quantified by comparing the signal from empty wells adjacent to bright ATP-containing wells against the true background.14Promega Corporation. How Sensitivity and Crosstalk Affect Your Bioluminescent Assay Results If you are seeing unexpectedly high background values in wells that should be empty or contain only media, crosstalk from adjacent wells is worth investigating. Using opaque white plates (which maximize signal capture) rather than clear-bottom plates can help, though it comes at the cost of not being able to image cells through the bottom of the plate before reading.
Practical Tips That Do Not Appear in the Package Insert
Experienced users accumulate a set of best practices that go beyond the manufacturer’s protocol. Equilibrating the reagent and plates to room temperature before combining them reduces well-to-well variability caused by temperature gradients. If you are running many plates in a session, stagger your reagent additions so each plate gets the same incubation time before reading, rather than adding reagent to all plates at once and reading them sequentially (which means the last plate sat longer than the first). Including a standard curve of known ATP concentrations on each plate, or at least on a representative plate in each batch, lets you convert relative light units to ATP amounts and catch any batch-to-batch shifts in reagent activity.
For drug screening, include both a no-treatment control (maximum viability) and a total-kill control (cells treated with a known toxic agent or detergent) on every plate. These define your 0% and 100% kill boundaries and allow you to calculate percent viability in a way that accounts for plate-specific background and maximum signal. When testing compounds that are dissolved in DMSO, your vehicle control should contain the same final concentration of DMSO as your treated wells, since DMSO itself becomes toxic to most cell types above about 0.5-1% final concentration.
Store reconstituted reagent at the temperature recommended by the manufacturer and avoid repeated freeze-thaw cycles, which degrade the luciferase enzyme and reduce signal over time. If you notice your positive control signal drifting downward across experiments, reagent degradation is the first thing to check. Many labs aliquot the reagent into single-use volumes immediately after reconstitution to avoid this problem entirely.