Why dilute in steps
Bacterial cultures, virus stocks and cell suspensions are often far too concentrated to count directly: a culture can hold a billion cells per milliliter. A single 1:1,000,000 dilution would need an impractically small transfer. A serial dilution gets there in steps, typically 1:10 at a time, with ordinary pipette volumes and much better accuracy.
The formulas
- Dilution factor per step = (transfer volume + diluent volume) ÷ transfer volume
- Total dilution = factor per step ^ number of steps
- Final concentration = starting concentration ÷ total dilution
- Expected colonies = final concentration × volume plated
- Back-calculation: CFU/mL = colonies ÷ (volume plated in mL × total dilution)
Worked example
A bacterial culture holds about 10⁸ CFU/mL. You transfer 100 µL into 900 µL of sterile broth, five times, then plate 100 µL of the last tube.
- Factor per step: (100 + 900) ÷ 100 = 10.
- Total dilution after 5 steps: 10⁵ = 100,000.
- Final concentration: 10⁸ ÷ 10⁵ = 1,000 CFU/mL.
- Colonies expected on the plate: 1,000 × 0.1 mL = 100, well within the countable range.
- Diluent needed: 5 × 900 µL = 4.5 mL.
Concentration at each step
| Tube | Total dilution | Concentration (from 10⁸/mL) | Colonies from 100 µL |
|---|---|---|---|
| 1 | 10¹ | 10⁷ | too many |
| 2 | 10² | 10⁶ | too many |
| 3 | 10³ | 10⁵ | too many |
| 4 | 10⁴ | 10⁴ | 1,000, too many |
| 5 | 10⁵ | 10³ | 100, countable |
| 6 | 10⁶ | 10² | 10, too few |
Plating two or three consecutive dilutions gives at least one plate in the 30 to 300 range.
Other common schemes
- 1:2 series, 500 µL into 500 µL: used for antibiotic MIC tests and standard curves.
- 1:5 series, 200 µL into 800 µL.
- 1:100 steps, 10 µL into 990 µL: fewer tubes but larger pipetting error.
Tips for accuracy
- Change tips at every step: carry-over from the more concentrated tube inflates the next one.
- Mix thoroughly before each transfer, by vortexing or pipetting up and down.
- Errors compound: a 2% pipetting error per step becomes about 10% after five steps.
- Spread or pour plates in duplicate and average the counts.