To evaluate whether the in vitro model (invasion and intracellular growth in Caco-2 cells) for determining virulence is a suitable alternative to the in vivo model (50% lethal dose), we compared the levels of virulence obtained with the two models. We tested L. monocytogenes strains isolated from food and clinical samples during three episodes of listeriosis occurring in Italy in the period 1993-1995. We also tested L. monocytogenes strains isolated from food during official control activities. The results obtained from the tested strains varied according to the experimental method adopted: the L. monocytogenes strains featuring the same genetic pattern showed a greater uniformity of response in vivo than in vitro. We can conclude that the in vitro model may be used as an alternative to the animal model to determine Listeria spp pathogenicity, though it cannot distinguish levels of virulence within the L. monocytogenes species.