Ninety-six isolates of presumptive or confirmed Listeria monocytogenes were obtained from local clinical (30 isolates) or food laboratories (66 isolates). Minimal biochemical analysis identified only 80% of these isolates as L. monocytogenes the remaining included L. seeligeri, 1%, or the non-haemolytic L. innocua, 19%. The 27 clinical and 50 food isolates, mainly from meat products, frozen confectionaries, and cheeses, confirmed as L. monocytogenes were compared biochemically and serologically. Twenty-one isolates, including some strains of L. innocua and L. seeligeri, were examined for pathogenicity in immunocompromized mice and 44 typed using bacterial restriction endonuclease DNA analysis (BRENDA). Only isolates of L. monocytogenes were found to be pathogenic. Biovar-typing of the isolates was unreliable and provided poor discrimination. Serogroups 1/2 and 4 predominated among clinical and food isolates and BRENDA provided better discrimination among isolates. Ten stable and reproducible restriction patterns were observed among the Listeria sp. isolates studied. Overall, a combination of techniques gave the best discrimination and indicated their potential for use as epidemiological tools.