Generation time in microbiology is the time it takes for a bacterial population to double, and calculating it comes down to a straightforward relationship between growth rate and the natural logarithm of 2. The core idea is simple: measure how fast cells are multiplying during their period of steady growth, then use that rate to figure out how long one doubling takes. What makes the calculation tricky in practice has less to do with the math and more to do with getting clean data from living organisms that do not always cooperate with tidy equations.
The Core Formula
Generation time (often written as g or td) is derived from the specific growth rate, μ. The growth rate itself is calculated from two measurements of population size taken at two different time points during exponential growth. In the most common version used in teaching and research labs, you take the natural logarithm of your population measurement at a later time, subtract the natural log of the earlier measurement, and divide by the elapsed time. That gives you μ. Then you divide ln 2 (which is roughly 0.693) by μ to get the doubling time.
In concrete terms: if your culture had an optical density of 0.1 at one hour and 0.4 at three hours, you would calculate μ as (ln 0.4 − ln 0.1) divided by (3 − 1), which works out to about 0.693 per hour. Dividing ln 2 by that growth rate gives a generation time of one hour. That approach, using optical density readings plugged into a line of best fit, is the standard spectrophotometric method used widely in microbiology labs.1MethodsX. An efficient, cost-effective method for determining the growth rate of sulfate-reducing bacteria using spectrophotometry
You will sometimes see the same relationship expressed with log base 2 instead of natural log, or with colony-forming units instead of optical density. The underlying logic does not change. What matters is that both measurements come from the same units and the same phase of growth.
Gathering the Raw Data
The formula only works if you feed it good numbers, and the way you get those numbers depends on the organism and the question you are asking. The two most common approaches in a standard lab are spectrophotometry and viable plate counts, and they measure fundamentally different things.
Spectrophotometry measures how cloudy (turbid) a liquid culture is. You shine light through a sample and record the optical density. As cells multiply, the culture gets cloudier and the OD reading climbs. This method is fast, cheap, and gives you a continuous curve you can sample at short intervals. The downside is that it measures total biomass, including dead cells and debris. If a large fraction of your population is dying while the rest divides, your OD curve will underestimate the true growth rate of the living cells.
Viable plate counts, on the other hand, measure only the cells that can form colonies on agar. You take a sample, dilute it through a series of tubes, spread it on plates, and count colonies after incubation. This gives you a more accurate picture of living, reproducing cells. Serial dilution-based growth curves can also be combined with other monitoring techniques to track both planktonic growth and biofilm formation simultaneously.2PubMed. Serial Dilution-Based Growth Curves and Growth Curve Synchronization for High-Resolution Time Series of Bacterial Biofilm Growth The tradeoff is that plate counts are labor-intensive, slower, and give you fewer time points to work with. If your sampling intervals are too far apart, you might miss the exponential phase entirely or catch only its tail end.
For most routine generation-time calculations in a teaching or research lab, spectrophotometry is the go-to method. You take readings every 15 to 30 minutes, plot them on a semi-log graph, identify the straight-line portion (that is your exponential phase), and pull your two data points from there. Plate counts become more important when you need to distinguish live cells from dead ones, or when you are working with organisms that grow in clumps or biofilms and do not give clean OD readings.
Why Exponential Phase Is the Only Phase That Counts
Bacteria in a fresh culture do not start dividing immediately. They go through a lag phase first, where cells are adjusting to their new environment, synthesizing enzymes, and gearing up for division. During this phase the population is essentially flat. If you pulled your data points from the lag phase, you would calculate a generation time of infinity, or something close to it, which tells you nothing useful.
After lag phase, the population enters exponential (or log) phase, where cells are dividing at a steady, maximal rate. This is the only window where the generation-time formula applies cleanly. The math assumes a constant rate of doubling, and that assumption holds only during exponential growth. Once nutrients start running low or waste products accumulate, the culture enters stationary phase and growth slows or stops. Data from stationary phase will give you an artificially long generation time that does not reflect the organism’s actual doubling capacity.
Identifying the boundaries of exponential phase is one of the trickiest parts of the whole exercise. On a semi-log plot (log of population size on the y-axis, time on the x-axis), exponential growth appears as a straight line. The start and end of that line are judgment calls, and different researchers drawing the line in slightly different places will get slightly different growth rates. Using a line of best fit through multiple data points, rather than relying on just two measurements, reduces this source of error considerably.
What Changes Generation Time
Generation time is not a fixed property stamped into an organism’s DNA. It shifts depending on the conditions the cells are growing in, sometimes dramatically.
Temperature
Temperature is the single biggest environmental lever on microbial growth rate. Every species has a range within which it can grow and an optimum where it grows fastest. E. coli, for example, can grow and divide across a wide band from about 23°C up to 40°C, and it can even tolerate pressures up to 400 atmospheres, though growth slows as conditions move away from the optimum.3PubMed Central. Pressure and temperature dependence of growth and morphology of Escherichia coli: experiments and stochastic model At its optimal temperature of about 37°C in rich media, E. coli can double every 20 minutes. Drop the temperature to 25°C and that generation time stretches to well over an hour. This sensitivity is exactly why food safety guidelines harp on keeping perishables out of the “danger zone” between roughly 4°C and 60°C.
Nutrients
The relationship between nutrient concentration and growth rate is one of the foundational tools of microbiology.4PubMed Central. Growth kinetics of suspended microbial cells: from single-substrate-controlled growth to mixed-substrate kinetics In rich media loaded with amino acids, sugars, and vitamins, bacteria grow at or near their maximum rate. In minimal media where cells have to synthesize their own building blocks, growth slows. In nature, where nutrient concentrations fluctuate constantly and organisms compete for limited resources, growth rates are almost always slower than in the lab.
Antimicrobial Stress
Even antibiotic concentrations below the level needed to kill bacteria outright can drag down their growth rate. Sub-lethal doses of antibiotics have been shown to reduce growth rates in veterinary pathogens, which has implications for how resistance can develop in low-exposure environments.5PubMed Central. Effects of sub-minimum inhibitory concentration antibiotic levels and temperature on growth kinetics and outer membrane protein expression in Mannheimia haemolytica and Haemophilus somnus When you calculate generation time under antimicrobial pressure, the number you get reflects a stressed organism, not its inherent maximum capacity. Comparing that stressed generation time with the unstressed value is actually one practical way to gauge how much an antimicrobial is affecting a pathogen’s fitness.
The Enormous Range of Microbial Generation Times
One of the more striking things about generation time as a metric is how wildly it varies across the microbial world. Under ideal lab conditions, E. coli famously doubles every 20 minutes. But that same bacterium, living in the gut of a mammalian host, doubles roughly every 15 hours. Vibrio cholerae in the wild has an estimated doubling time of about 1.1 hours, while Salmonella enterica in natural settings can take around 25 hours to double. Across the board, generation times measured in wild populations are longer than those measured in the laboratory.6PubMed Central. The distribution of bacterial doubling times in the wild
At the extreme slow end, microorganisms living in deep-sea sediments, hundreds of meters below the ocean floor, subsist on vanishingly small amounts of energy. Their estimated generation times stretch into tens, hundreds, and possibly thousands of years.7PubMed. Slow Microbial Life in the Seabed Calculating generation time for those organisms requires entirely different methods than a lab growth curve, and the numbers carry large uncertainties, but they illustrate that microbial “growth” spans an almost incomprehensible range of timescales.
Researchers have compiled databases of minimal generation times for hundreds of bacterial and archaeal species, drawing from published primary literature, which helps establish the baseline speed of different lineages and even correlates with genomic features like codon usage and ribosomal RNA copy number.8PLoS Genetics. The Systemic Imprint of Growth and Its Uses in Ecological (Meta)Genomics In other words, fast-growing bacteria tend to carry certain genomic signatures that distinguish them from slow growers. This means generation time is not just a physiological measurement; it is connected to the evolutionary strategy of the organism.
Beyond the Standard Growth Curve
The classic batch-culture approach described above, where you inoculate a flask, track the population over time, and extract your growth rate from the exponential portion, works well for fast-growing lab organisms in controlled conditions. But microbiology has increasingly moved toward questions that this method cannot answer. How fast is a single cell dividing? How fast are unculturable species growing in soil or ocean sediments? What is the generation time of a pathogen inside a living host? Answering those questions requires different tools.
Single-Cell Imaging
Microfluidic devices, sometimes called “mother machines,” trap individual bacteria in tiny channels and image them continuously under a microscope. Software then tracks each cell’s elongation and division over many generations. This approach produces a direct measurement of generation time for individual cells, not a population average, revealing the cell-to-cell variation that bulk cultures smooth over.9eLife. Tools and methods for high-throughput single-cell imaging with the mother machine That variation turns out to be substantial. Even in a genetically identical population growing under constant conditions, some cells divide faster than others, and a fraction may not divide at all.
Isotopic Probing for Environmental Microbes
For organisms that cannot be cultured in the lab, which represents the vast majority of microbial diversity on Earth, generation time has to be inferred by other means. One powerful approach uses water labeled with a heavy oxygen isotope. Actively growing cells incorporate the heavy oxygen into their newly synthesized DNA. By separating DNA by density and sequencing it, researchers can figure out which specific taxa were growing and how fast, even in a complex soil community with thousands of species present.10Soil Science Society of America Journal. Quantitative stable isotope probing with H218O to measure taxon‐specific microbial growth This method has opened up growth-rate measurements for organisms that were essentially invisible to traditional cultivation.
Common Pitfalls That Throw Off Your Numbers
If you have ever run a generation-time calculation and gotten a number that looked suspiciously fast or absurdly slow, chances are good that one of a few common errors was at work.
- Sampling outside exponential phase: Pulling your data from lag phase or early stationary phase inflates the apparent generation time. On a semi-log plot, look for the steepest straight-line segment and use only that window.
- Too few time points: With only two or three measurements, you cannot reliably identify where exponential phase begins and ends. More data points let you fit a better line and reduce the impact of any single noisy reading.
- Ignoring cell clumping: Organisms that form chains, clusters, or biofilms will give misleadingly low OD readings relative to their actual cell count, because clumped cells scatter light differently than dispersed ones. If your organism clumps, plate counts or alternative methods may be more reliable.
- Confusing population heterogeneity with a clean average: Even in a clonal population of Salmonella, individual cells show highly diverse growth dynamics when tracked by time-lapse microscopy. Some cells filament or lyse before dividing, while others that do grow form microcolonies at very different rates.11PubMed Central. Stochasticity in colonial growth dynamics of individual bacterial cells The generation time you calculate from a bulk culture is a population average that masks this underlying stochasticity. That average is useful, but it does not describe every cell in the flask.
- Mixing up log bases: The formula works with natural logs (ln) or with log base 2 or log base 10, but you have to be consistent. If you use ln for the population measurements, you divide by ln 2 (0.693) to get generation time. If you use log10, you divide by log10 2 (0.301). Mixing bases silently produces a number that is off by a factor of about 2.3.
Practical Applications of Generation Time
Generation time might sound like a purely academic exercise, something you calculate in a microbiology lab course and never think about again. But it has real consequences in several fields.
In food safety, the generation time of foodborne pathogens at different temperatures is the backbone of predictive models that determine shelf life and safety guidelines. These models take real temperature profiles from food processing and storage, feed them through growth equations calibrated with generation-time data, and predict whether a batch of food is likely to reach a dangerous level of contamination before its expiration date.12PubMed Central. Quantitative microbiology: a basis for food safety The “keep hot food above 60°C, cold food below 4°C” rule exists because the generation times of most dangerous bacteria become long enough at those temperatures that populations cannot grow to harmful levels within typical holding times.
In clinical microbiology, generation time determines how quickly an infection can escalate and influences how antibiotics are dosed. A pathogen with a 20-minute generation time can go from a handful of cells to millions in under seven hours. Slower-growing organisms like Mycobacterium tuberculosis, with a generation time of roughly 15 to 20 hours, take weeks to form visible colonies on culture plates, which is why TB diagnosis and drug-susceptibility testing historically required so much patience.
In industrial biotechnology, generation time directly affects the economics of producing anything from insulin to biofuels. A production strain that doubles every 30 minutes instead of every 60 minutes cuts fermentation time roughly in half for the same yield. Engineers spend considerable effort optimizing growth conditions and selecting fast-growing strains precisely because shaving minutes off the generation time scales up into significant cost savings at production volumes.
The Gap Between Lab Numbers and Reality
One thing worth internalizing is that the generation time you calculate from a lab growth curve is a best-case scenario. It tells you how fast an organism can divide when it has everything it needs: optimal temperature, abundant nutrients, no competitors, no predators, no immune system attacking it. In nature, those conditions essentially never exist simultaneously.
The gap between lab and wild generation times is not a small discrepancy. For E. coli, the difference is roughly 45-fold: 20 minutes in the lab versus about 15 hours in the wild.6PubMed Central. The distribution of bacterial doubling times in the wild For some species the gap is narrower, for others wider, but it exists for every organism that has been studied in both contexts. This matters because it means lab-derived generation times, while useful as benchmarks, can be seriously misleading if you use them to predict how fast bacteria are actually multiplying in a patient, a river, or a slice of cheese sitting on a counter.
Environmental microbiologists have increasingly recognized that measuring growth rates in situ, rather than extrapolating from lab cultures, is essential for understanding microbial ecology. Techniques like the isotopic probing method described earlier, along with metatranscriptomics and direct microscopic observation, are filling in a picture that lab growth curves alone could never provide. The generation time you calculate in a teaching lab is a starting point for understanding microbial growth, not the final word on what any given bacterium is actually doing in the environments where it lives.