Aim: To determine growth initiation differences of Listeria monocytogenes between a cheesemaking context, milk and tryptic soy broth (TSB). Methods and Results: A laboratory-scale cheese was made with a mix of two strains of L. monocytogenes at four initial pH values, five water activity (a(w)) values and two contamination levels at 30 degrees C. Counts of L. monocytogenes were determined at time 0 and after 8 h of cheese manufacture. Milk and TSB at the same pH and a(w) conditions were inoculated with the L. monocytogenes mix in multi-well plates. Growth was determined by plating each well onto Agosti & Ottaviani Listeria Agar after 8 h of incubation at 30 degrees C. Each condition was repeated six times, and growth initiation probability was modelled with logistic regression models. Growth initiation boundaries were obtained for each matrix type. The results showed that the growth limits were matrix dependent. In the three matrix types, a(w) was the most important factor affecting the probability of growth initiation. Contamination level affected growth TSB and cheesemaking conditions. Conclusions: The interface wideness and position in cheese, milk and TSB were dissimilar, indicating that the use of models evaluated in TSB or milk could not be used to predict the behaviour of L. monocytogenes under cheesemaking conditions. Significance and Impact of the Study: Predictive models generated in liquid media are not necessarily adaptable to solid food, and the generation of real food models is necessary.