Clostridium perfringens is one of the five major food borne pathogens in the western world (expressed in cases per year). Symptoms are caused by an enterotoxin, for which 6% of type A strains carry the structural gene. This enterotoxin is released when ingested cells sporulate in the small intestine.Research on C. perfringens has been limited to a couple of strains that sporulate well in Duncan and Strong (DS) medium. These abundantly sporulating strains in vitro are not necessarily a representation of the most dangerous strains in vivo. Therefore, sporulation was optimized for C. perfringens strains in general. None of the tested media and methods performed well for all strains, but Peptone-Bile-Theophylline medium (with and without starch) yielded highest spore numbers. DS medium may be optimized by adding theophylline, but PBT(S) medium is most suitable.Since agar media influence cell counts of C. perfringens, six different media for enumerating C. perfringens from foods were tested, four of which were analyzed in an international collaborative trial. The examined media were equal from a microbiological point of view, but Tryptose-Sulphite-Cycloserine (TSC) agar was most favorable due to ease of use.C. perfringens outbreaks are often associated with preparation of large food quantities. Large volumes of food are difficult to cool down, and cooling is one of the most important steps in preventing growth of C. perfringens. Data from the Dutch Inspectorate for Health Protection and Veterinary Public Health showed that pea soup was relatively often contaminated with C. perfringens. Therefore, the effect of different cooling procedures on growth of this pathogen was tested in pea soup. It was shown that cooling requirements stated by Dutch law allowed more than 1 log cycle growth of C. perfringens. This level of cell increase is inadmissible according to USA law. During subsequent refrigerated storage cell numbers remained constant. Therefore, a model was designed to predict the effect of various cooling scenarios on growth of this pathogen in pea soup. This model can be used to design Good Manufacturing Practices (GMP) procedures.It can be concluded that C. perfringens is a heterogeneous species. To predict worst-case scenarios of the behavior of this pathogen in food, research should not focus on a couple of strains. To limit excessive screening of strains, a gamma-type growth model, based on general and interpretable parameters, was shown to accurately predict growth of C. perfringens. Combined with models that describe food-processing procedures, the growth model can serve as a good tool for GMP.