Written and maintained by CASRAI Editorial Board
Last updated
To count colonies on an agar plate, pick the plate whose colony number falls inside the countable range for your method, count every distinct colony (using a colony counter with a grid if you have one), average the replicate plates, then divide by the dilution plated and the volume plated to get CFU per mL. The countable range is not the same everywhere: the textbook rule is 30 to 300 colonies, but FDA’s Bacteriological Analytical Manual changed its range to 15 to 300 in March 2025, so use the range written in your own method.
This guide covers general bench counting of bacterial colonies on agar plate cultures after dilution and plating. It does not replace the validated method or SOP your laboratory works under.
What you need
- Incubated plates from a dilution series, ideally with replicate plates at each dilution (see serial dilution technique for how the series is built).
- A colony counter. The FDA method lists a dark-field, Quebec or equivalent counter with a suitable light source and a grid plate.
- A tally counter or marker, and a calculator.
- Your method’s countable range, the dilution of each plate and the volume plated on each plate. Plates are normally poured or spread in Petri dishes.
Steps
- Find the countable plate. Look across the dilution series and choose the plates whose colony number lies inside the range your method specifies. FDA BAM Chapter 3 (aerobic plate count) uses 15 to 300 colonies per plate; the same chapter notes that for pour plates USDA and AOAC use 30 to 300 and APHA uses 25 to 250. A USDA veterinary biologics SOP counts plates with 30 to 300 colonies and says not to use plates with fewer than 30 colonies as estimates unless the firm’s own procedure allows it.
- Count each distinct colony. Place the plate on the counter, use the grid to keep your place, and count every separate colony once. Colonies that arise as a spreading film or chain follow the spreader rules below rather than being counted growth by growth.
- Handle spreaders. FDA describes three kinds: a chain of colonies from a broken-up clump, growth in a water film between the agar and the dish bottom, and growth in a water film at the edge or surface of the agar. If spreaders (including repressed growth) cover more than 50% of the plate, or repressed growth alone covers more than 25%, report the plate as a spreader plate. Otherwise count each of the three spreader types as one source: a single chain is one colony, chains from separate sources are one colony per source, and each individual growth within a chain is not counted separately. The other two types usually appear as distinct colonies and are counted as such. Add the spreader count to the colony count.
- Average the replicates. Where you plated replicates at the countable dilution, take the mean count. The USDA SOP’s own worked example averages three plates (82, 79 and 88).
- Convert to CFU per mL. The USDA SOP gives the calculation in words: the average count divided by the dilution plated multiplied by the mL plated equals CFU/mL. Illustrative arithmetic, not a source figure: a plate from the 10-6 dilution with 150 colonies, 1 mL plated, gives 150 / (10-6 x 1) = 1.5 x 108 CFU/mL.
- Round and report. FDA’s method reports only the first two significant digits and rounds only at the final conversion: a third digit of 6 to 9 rounds up, 1 to 4 rounds down, and a 5 rounds up if the second digit is odd and down if it is even (12,700 becomes 13,000; 15,500 becomes 16,000; 14,500 becomes 14,000). Counts outside the countable range are reported as estimated.
Counting bacterial colonies on agar plate cultures: which range to use
| Source | Countable colonies per plate |
|---|---|
| FDA BAM Chapter 3 (January 2026 edition, adopting the ISO range) | 15 to 300 |
| USDA and AOAC (as cited in FDA BAM Chapter 3) | 30 to 300 |
| APHA (as cited in FDA BAM Chapter 3) | 25 to 250 |
| USDA APHIS CVB-SOP-0104.03 Standard Bacterial Plate Count (15 March 2024) | 30 to 300 |
The ranges differ because they belong to different methods. A 2016 Journal of Bacteriology editorial, “Classic Spotlight: Plate Counting You Can Count On,” notes that most textbooks and lab manuals recommend plates yielding roughly 30 to 300 colonies, because with too few colonies the count may not be accurate and with too many it is hard to tell individual colonies apart. The 1916 study it revisits actually found plates with roughly 40 to 200 colonies the most reproducible, so the range is a convention that methods set for themselves.
What can go wrong
Too few colonies
A low count may not be accurate, and the USDA SOP does not use plates under 30 colonies for CFU estimates unless the firm’s procedure says otherwise. If no dilution shows any colonies, the FDA method reports the result as less than 1 times the lowest dilution used, marked with an asterisk as an estimate.
Too many colonies
Crowded plates make individual colonies hard to distinguish. Counts above the range are reported as estimated; FDA’s method handles too-numerous-to-count plates with a maximum-likelihood calculation rather than a guessed number.
Spreaders and touching colonies
Use the spreader rules in step 3. Do not count every growth in a chain as its own colony.
Uneven or pooled inoculum
The USDA SOP requires the inoculum to be spread evenly over the agar surface and not allowed to pool around the edges, and calls for a fresh spreader for each dilution plated. Pooled or carried-over inoculum distorts the colony pattern and the count. That SOP also has plates held in a 35 +/- 2 C incubator overnight or dried in a biosafety cabinet before use, and says to avoid spreading inoculum right to the plate edge.
Applying the wrong rounding or range
A count that is valid under one method’s range may be an estimate under another. Check which range and rounding rules your method names before you report.
When to get professional help
Ask the person who owns your laboratory’s method, your quality assurance lead or your biosafety officer when counts repeatedly fall outside the range, replicate plates disagree widely, you suspect contamination, or you are unsure which method’s range applies to a regulated or release test. This guide gives general counting practice only, not individual medical or regulatory advice.
FAQ
What is a good number of colonies to count on a plate?
One that falls in the range your method specifies: 30 to 300 under the USDA SOP and in textbook practice, 25 to 250 under APHA, and 15 to 300 under FDA BAM since March 2025.
What is a CFU?
CFU stands for colony-forming unit. Plate counts are reported as CFU per mL (or per dose) because one visible colony may come from a single cell or from a clump or chain of cells, so the count is of colony-forming units rather than individual cells.
Why is the usual range 30 to 300?
Below the range the count is less reliable, and above it colonies crowd together and cannot be told apart. Different regulators and standards set different limits within that logic, which is why the FDA, USDA and APHA ranges differ.
What do I do when colonies touch or spread?
Apply the spreader rules in step 3: count a chain from one source as one colony, and report the plate as a spreader plate when spreaders cover more than 50% of it.
Can a machine count them for me?
Colony counters with a grid and light source help a person count; see also automated cell counters for the related instrument category. Counting rules for automated spiral plating differ from the manual rules here.
Where this fits
Colony counts feed susceptibility and kill-rate work such as minimum inhibitory concentration testing and time-kill assays. For the related instrument category, see viable air samplers.
Sources
- U.S. FDA, Bacteriological Analytical Manual (BAM) Chapter 3: Aerobic Plate Count, January 2026 edition, https://www.fda.gov/food/laboratory-methods-food/bam-chapter-3-aerobic-plate-count. Checked 29 October 2026.
- USDA APHIS Center for Veterinary Biologics, CVB-SOP-0104.03, Standard Bacterial Plate Count, 15 March 2024, https://usbiotechnologyregulation.mrp.usda.gov/sites/default/files/cvb-sop-0104.pdf. Checked 29 October 2026.
- “Classic Spotlight: Plate Counting You Can Count On,” Journal of Bacteriology 198:3127, 2016, https://pmc.ncbi.nlm.nih.gov/articles/PMC5105891. Checked 29 October 2026.
