From inoculation experiments in laboratory media, predictive microbiology (PM) derives equations to quantitatively describe the behavior of microorganisms in foods depending on intrinsic and extrinsic factors (controlling factors). Meanwhile, numerous growth, survival, and thermal inactivation (death) models have been elaborated for the most important foodborne pathogens. The Food MicroModel and the Pathogen Modeling Program are available as user-friendly software applications. Although all PM models are simplifications of biological mechanisms and the models actually available still have their limitations, comparisons with independent data from the literature indicate that predictions of most models are in the worst case fail-safe and that their systematic errors do not exceed those of inoculated pack experiments. Once a model has been validated for a specific type of food, it can be applied at all stages of food production and distribution. PM models are already used to conduct HACCP studies and are powerful tools for microbiological risk assessment in particular. Under both aspects, there are various applications for PM: determining necessary time-temperature combinations during heating processes, estimating the risk of pathogen growth during planned storage time and conditions, and examining new formulations for potential microbiological hazards. Traditionally, these tasks were done by using microbiological challenge testing (MCT, inoculated pack experiments). However, these experiments are expensive and time consuming. Moreover, their results are only valid for the product being tested and the conditions of its processing and storage. If changes are planned or occur occasionally, new MCTs have to be conducted under changed conditions. PM models are most effectively used at the stage of product development. As they allow a fast first estimation about the behavior of microorganisms, their application enables food microbiologists to recognize whether the product has to be modified in formulation or process without losing time waiting for MCT results, thus avoiding unacceptable levels of risk for consumers.