CFU GUIDE

How to calculate CFU/mL from colony counts

Step-by-step CFU/mL calculation from colony counts, dilution fraction, and plated volume, with worked examples, countable-range guidance, and common mistakes.

A raw colony count is not a concentration. It becomes one only when you combine it with two more numbers: the dilution of the tube you plated, and the volume of that tube that went onto the plate. This guide walks through the full calculation the same way the CFU calculator does, then covers the judgement calls — which plate to count, how replicates are handled, and where the arithmetic goes wrong.

The formula

For a liquid sample:

CFU/mL = colonies counted ÷ (dilution fraction × plated volume in mL)

Three terms carry the whole calculation:

• Colonies counted — the colonies on the plate, or the pooled estimate from eligible replicate plates.
• Dilution fraction — the portion of the original sample present in the tube you plated. A 10⁻³ dilution is a fraction of 0.001. The same dilution stated as a “dilution factor” of 1,000 is the reciprocal of the fraction; the two conventions describe one dilution and must not be mixed in one equation.
• Plated volume — the volume of that dilution spread or poured onto the plate, in millilitres.

For a solid sample, the mass and the homogenization volume join the formula:

CFU/g = colonies × homogenization volume ÷ (sample mass × dilution fraction × plated volume)

The CFU calculator applies whichever form matches the sample type you select and shows every substituted value.

Worked example

You count 86 colonies on a plate spread with 0.1 mL taken from a 10⁻³ dilution tube.

1. State the dilution as a fraction: 10⁻³ = 0.001.
2. Work out how much of the original sample sits on the plate: 0.001 × 0.1 mL = 10⁻⁴ mL.
3. Divide the count by that amount: 86 ÷ 10⁻⁴ = 8.6 × 10⁵ CFU/mL.

The same calculation written with the dilution factor would be CFU/mL = colonies × dilution factor ÷ plated volume = 86 × 1,000 ÷ 0.1. Both forms give the same answer, but the fraction form above is the safer habit, because mixing the two conventions in one equation is the classic route to a six-order-of-magnitude error.

Choosing which plate to count

Most protocols treat roughly 30–300 colonies as countable. Some methods use different windows: 25–250 for certain FDA BAM and USP procedures, 10–150 for yeasts and moulds. The window exists for two reasons. Below it, statistical noise in a small count dominates — nine colonies and fifteen colonies are not reliably different. Above it, colonies begin touching and competing for nutrients, so some fail to appear and the count reads low.

Plates below the window are flagged TFTC (too few to count); plates above it are TNTC (too many to count). The calculator lets you set the window your method specifies and omits out-of-range plates from pooled averages.

Replicate plates

If two plates from the same dilution are both countable, pool them rather than averaging the concentrations:

pooled CFU/mL = Σcolonies ÷ Σ(dilution fraction × plated volume)

This is the estimator the calculator uses when you confirm that the entered plates are replicates of one specimen. It weights each plate by its own effective volume, which is the statistically sound treatment when volumes differ.

Pour plate or spread plate

The mechanics differ — a pour plate is mixed into molten agar, a spread plate sits on the surface — but the arithmetic does not. Both count colonies that grew from the volume you delivered, so the plated volume in the formula is the volume actually transferred to the plate. One practical difference: spread plates generally cannot take much more than about 0.1–0.3 mL without pooling issues, while pour plates commonly receive 1 mL. If you mix a sample with agar before pouring, remember that the agar itself is not a dilution — only added diluent is.

Common mistakes

• Confusing factor and fraction. Dividing by 1,000 where 0.001 belongs (or the reverse) shifts the answer by six orders of magnitude. Pick one convention and check the magnitude of your answer.
• Forgetting the plated volume. When you plate 0.1 mL instead of 1 mL, the result is ten times higher than the count alone suggests. The formula only works when V is in millilitres and matches what you actually dispensed.
• Missing a preparation dilution. If the sample was diluted during preparation — a 1 g aliquot into 9 mL, for example — that step is part of the chain and must be included, as in the solid-sample formula.
• Using the wrong tube’s dilution. Write the dilution on every tube. Scaling a count from the 10⁻² plate by the 10⁻³ fraction is a classic transcription error.

Sanity-checking the result

If you plated two adjacent dilutions, the counts should differ by roughly the step ratio — about tenfold for a 10-fold series. A 10⁻³ plate with 86 colonies and a 10⁻² plate with 90 colonies is not plausible; one of the labels or calculations is wrong. Also sanity-check the magnitude: a result like 8.6 × 10⁵ CFU/mL is typical of an overnight bacterial culture; 8.6 × 10¹⁵ should make you re-check your exponents.

Limitations

One colony can come from a single cell or from a cluster that stayed together, so the result estimates viable, culturable units under your incubation conditions — not a direct cell count. Cells injured by processing, cells in a viable-but-nonculturable state, and cells that simply dislike your medium never appear. A zero-colony plate means nothing grew under those conditions; it does not prove the sample is sterile. Report the result with its basis (CFU/mL or CFU/g), the medium, incubation conditions, and the dilution that produced the countable plate.

Open the CFU calculator to run your own counts, or continue with CFU/mL vs CFU/g for solid-sample preparation and what one CFU represents for interpretation.

Open the CFU calculator ↗