MTT Assay: Cell Viability Measurement and Pitfalls

The MTT assay is a colorimetric plate-based test that estimates cell viability from metabolic activity rather than by counting individual cells. A yellow tetrazolium salt — 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, universally shortened to MTT — is reduced by metabolically active cells into an insoluble purple formazan product; after solubilization, the amount of formazan is read as absorbance on a plate reader and used as a proxy for how many cells in the well are alive and metabolically functioning. First described by Tim Mosmann in 1983, it remains one of the most widely used cytotoxicity and proliferation assays because it is cheap, scales to a 96-well format, and needs no specialized instrument beyond a plate reader. It is also, as the pitfalls section below covers in detail, a proxy with real blind spots — not a direct measurement of live cell number.

What the MTT assay actually measures

MTT reduction depends on NAD(P)H-linked reducing activity inside the cell — the exact enzymology is still not fully characterized, but the reaction is generally attributed to NADH- and NADPH-dependent oxidoreductase activity rather than a single dedicated mitochondrial enzyme. In practice this means the assay reports reducing/metabolic capacity per well, which correlates with live cell number under normal culture conditions but is not the same thing as live cell number. A treatment that slows a cell’s metabolism or glycolytic rate without killing it can lower the MTT signal even though the cells are still alive; a treatment that leaves cell number unchanged but increases metabolic rate can raise it. This is the central fact to keep in mind when interpreting an MTT result: it is a metabolic-activity readout across a population of cells in a well, not a cell count.

The chemistry: tetrazolium reduction to formazan

MTT is water-soluble and yellow in its oxidized form. Inside metabolically active cells, NAD(P)H-dependent reducing activity cleaves the tetrazolium ring and converts MTT into formazan — a purple product that is insoluble in aqueous culture medium and accumulates as needle-like crystals inside (and sometimes outside) the cell. Because formazan is insoluble, it cannot be read directly in the culture medium; it has to be dissolved first. Standard solubilizing agents are DMSO, acidified isopropanol, or an SDS-detergent solution, each added directly to the well after the culture medium is removed (or, for the acid-isopropanol method, after aspirating supernatant). Once dissolved, the purple solution is read at 570 nm absorbance, typically with a reference wavelength around 630–690 nm subtracted to correct for plate/plastic background — consult your specific kit or protocol for the exact reference wavelength it specifies, since this varies slightly by formulation.

Standard workflow

• Seeding. Cells are seeded in a 96-well plate at a density appropriate to the cell line and the planned incubation window — dense enough to give a measurable signal, sparse enough that control (untreated) wells don’t reach confluence and plateau before the assay endpoint, which would flatten out real differences between treatment conditions. See cell culture reference numbers for typical seeding-density starting points by vessel format.
• Treatment. Test compound (or vehicle control) is added, usually after an overnight attachment period for adherent lines, and incubated for the exposure window the experiment calls for — anywhere from a few hours to several days depending on the biology being tested.
• MTT incubation. MTT solution (commonly 0.2–0.5 mg/mL final concentration) is added directly to the wells and incubated, typically 1–4 hours at 37°C, until purple formazan crystals are visible under the microscope. Longer incubation increases signal but also increases the chance of crystal loss during medium removal and interference from non-cellular reduction — incubation time should be fixed and identical across every plate in a comparison, not varied to “get a better signal” on a weak plate.
• Solubilization. Medium is removed, the solubilizing agent is added, and the plate is mixed (often on a shaker, in the dark) until the crystals are fully dissolved and the color is uniform well-to-well.
• Reading. Absorbance is read at 570 nm on a plate reader, with a reference wavelength subtracted per protocol.

Controls and blanks — don’t skip these

Every MTT plate needs, at minimum: an untreated (or vehicle-only) control to set the 100% viability baseline; and blank wells — medium, MTT, and solubilization solvent, with no cells — to measure and subtract background absorbance from the medium, plasticware, and phenol red itself. Skipping the blank is a common source of an artificially inflated viability reading, especially in phenol-red-containing medium, which itself absorbs at wavelengths close to the formazan read. Where the test compound is colored or itself absorbs near 570 nm, add a compound-only, no-cell well as well, so its contribution can be subtracted rather than misread as formazan signal.

Common pitfalls and sources of error

• Direct chemical interference from the test compound. Some compounds — or their metabolites — reduce MTT directly, independent of cellular metabolism, or interfere with formazan formation or solubility. This can over- or under-estimate apparent viability regardless of what’s actually happening to the cells. A cell-free well containing compound, medium, and MTT (no cells) is the standard control for catching this; if that well develops color, the compound is interfering directly and the assay result for that compound needs a different viability method or a correction.
• Reducing agents and antioxidants in the medium or treatment panel. Ascorbic acid and other reducing/antioxidant compounds can reduce MTT non-enzymatically, inflating the signal in wells where they’re present.
• Phenol red and serum background. Phenol red absorbs in a range close to the formazan read and can shift results if blank subtraction isn’t done consistently; phenol-red-free medium removes the issue entirely where the cell line tolerates it.
• Edge effects. Outer wells of a 96-well plate are more exposed to evaporation and temperature gradients than inner wells, which skews their readings independent of the actual treatment. The standard fixes are to fill unused outer wells with sterile PBS or water (a “moat” plate) rather than leaving them empty or using them for data, and to randomize where treatment conditions land on the plate rather than always using the same columns.
• Incomplete or inconsistent solubilization. If formazan crystals aren’t fully dissolved before reading, absorbance reads low and unevenly across the plate, in a pattern that can look like a treatment effect but is really a mixing artifact. Confirm crystals are visibly dissolved and the color is uniform before reading.
• Timing drift across a batch of plates. Absorbance keeps rising the longer MTT incubates, so plates read at slightly different times after MTT addition aren’t directly comparable. Process plates in a fixed, repeatable order and timing, the same discipline that matters for trypan blue counting on a hemocytometer (see below) for the same underlying reason — a time-dependent measurement drifts if it isn’t standardized.
• Confusing metabolic signal with cell number. Because MTT reports metabolic/reducing activity rather than a direct count, a compound that is cytostatic (slows or stops proliferation without killing cells) and a compound that is genuinely cytotoxic (kills cells) can produce a similar drop in MTT signal. Where that distinction matters to the experiment’s conclusion, pair MTT with a direct cell-count method or a live/dead-specific assay rather than relying on the metabolic proxy alone.

MTT vs. hemocytometer counting: two different definitions of “viability”

MTT and manual hemocytometer cell counting with trypan blue are both called “viability” methods, but they measure genuinely different things and answer different questions:

• Hemocytometer + trypan blue is a direct, per-cell count: a technician looks at individual cells under a microscope and classifies each one as membrane-intact (excludes the dye, scored live) or membrane-compromised (takes up the dye, scored dead). It reports an actual number of live and dead cells in a known sample volume, but only for the one sample being counted at that moment, and only based on membrane integrity — as the hemocytometer guide covers, an early-apoptotic cell with an intact membrane will score as “viable” by trypan blue even though it’s already committed to dying.
• MTT is an indirect, population-level metabolic readout: it never looks at or counts an individual cell. It reports total reducing activity across everything in the well, read as absorbance, and infers a viability trend from that. It doesn’t tell you how many cells are in the well, doesn’t distinguish “fewer cells” from “the same number of cells with lower metabolic activity,” and is vulnerable to chemical interference from whatever else is in the well — none of which are failure modes a hemocytometer count has, because a person is directly looking at cells rather than reading a proxy signal.

In practice, the two are complementary rather than substitutes. A hemocytometer count is the right tool for establishing a starting cell number before an experiment (seeding density) or for a low-throughput, high-confidence viability check on one sample. MTT (and other plate-based colorimetric or luminescent assays) is the right tool for screening many conditions or compound concentrations in parallel — a dose-response cytotoxicity screen across a 96-well plate is exactly the throughput case a hemocytometer can’t practically cover, but MTT can. Neither is a universally “better” viability measurement; they answer different questions at different scales, and a rigorous cytotoxicity study often uses both — an MTT screen to identify a concentration range, followed by targeted hemocytometer or flow-cytometry-based confirmation (see the Annexin V/PI apoptosis assay for a method that, unlike either MTT or trypan blue, can distinguish live, early-apoptotic, late-apoptotic, and necrotic cells directly) on the conditions that matter most.

Newer alternatives, briefly

MTT is not the only tetrazolium-based option. MTS uses a related tetrazolium compound together with an electron-coupling reagent (commonly PMS or PES) and is reduced directly into a water-soluble formazan — it skips the solubilization step MTT requires, at the cost of a somewhat higher plate background. WST-1/WST-8 assays work on the same water-soluble-formazan principle. Further from the tetrazolium family, ATP-based luminescent assays (commercially, CellTiter-Glo is the best-known) lyse cells directly and measure ATP via a luciferase reaction, producing a stable luminescent signal within minutes with no incubation period needed for the readout itself; they are generally reported as more sensitive and less prone to some of the direct-chemical-interference issues MTT has, though they carry their own pitfalls (compounds that affect ATP metabolism or interfere with luciferase chemistry directly, for instance). None of these displaces MTT outright — it remains a reasonable, low-cost default for a standard cytotoxicity screen, with the tradeoffs above worth knowing before choosing it over an alternative for a specific assay.

Frequently asked questions

What does the MTT assay actually measure?

Metabolic reducing activity across a population of cells in a well, used as a proxy for viable cell number — not a direct count of live cells, and not specific to any single enzyme or organelle.

What wavelength do I read an MTT assay at?

570 nm absorbance is standard, typically with a reference wavelength around 630–690 nm subtracted for background — check the exact reference wavelength against the protocol or kit you’re using, since it varies slightly by formulation.

Why is my MTT data inconsistent between plates or replicates?

The most common causes are inconsistent MTT incubation timing, incomplete formazan solubilization, missing or inconsistent blank-well subtraction, and edge effects from unused or unfilled outer wells. Standardizing timing, confirming full solubilization by eye, and using a consistent blank/control layout resolves most of it.

Can a test compound interfere with the MTT reading directly?

Yes — some compounds reduce MTT directly or interfere with formazan solubility independent of any effect on the cells. A cell-free well with compound, medium, and MTT (no cells) is the standard way to check for this before trusting the result.

Is MTT the same as trypan blue exclusion counting?

No. MTT is an indirect, plate-based metabolic-activity readout across a whole well; trypan blue with a hemocytometer is a direct, per-cell microscopic count based on membrane integrity. They measure different things and have different blind spots — see the comparison above.

What’s the difference between MTT and MTS?

Both are tetrazolium-based, but MTS produces a water-soluble formazan directly, so it skips MTT’s separate solubilization step, at the cost of a higher plate background.