A different result was reported by Silbernagel and Linderberg (2003), who in a study on artificial contamination with Proteus vulgaris at different inoculum levels in milk samples evaluated the performance of the Petrifilm EB system and compared the results to those obtained with the plate count method and the Most Probable Number (MPN) technique. In this study, the results showed that there was no statistical difference between the methods at an medium level of inoculation (103 CFU mL-1). However, as for low (102 CFU mL-1) and high inoculation concentrations (104 CFU mL-1) they did exhibit a statistical difference, with the Petrifilm EB system presenting the highest mean value.