Three different Listeria monocytogenes strains, LO28 (a laboratory strain with truncated InlA), 4446 (a clinical isolate) and 7291 (a food isolate), were compared in a guinea-pig model designed to mimic food-borne exposure. The objectives were (1) to verify the applicability of the animal model for distinguishing between Listeria with different virulence properties and (2) to explore whether it was possible to reduce the required number of animals by dosing with mixed cultures instead of monocultures. Consistent with in vitro observations of infectivity in Caco-2 cells, faecal densities and presence in selected organs were considerably lower for LO28 than for the other two strains. Additionally, the animal study revealed a difference in prevalence in faeces as well as in internal organs between the clinical isolate and the food isolate, which was not reproduced in vitro. Dosage with monocultures of Listeria strains gave similar results as dosage with a mixture of the three strains; thus, the mixed infection approach was a feasible way to reduce the number of animals needed for determination of listerial virulence.