Cheese potentially allowing the growth of Listeria monocytogenes must be free of the pathogen in 25 g before being put on the market, while 100 cfu/g is tolerated when the pathogen is unable to grow. Challenge tests were performed in order to assess the growth potential of L. monocytogenes in at least one batch of 32 Belgian cheese varieties from 32 factories. All varieties were grouped in four categories: unripened acid-curd cheeses, mold-ripened soft cheeses, smear-ripened soft cheeses and ripened semi-hard cheeses. Associated microflora and cheese physicochemical characteristics were also studied. A cocktail of three strains was used to inoculate cheese on the first day of shelf-life, and samples were stored until the end of shelf-life at 7-9 °C. Growth potential was considered as the difference (a) between median contamination at the end and at the beginning of the test or (b) between the highest value at the end of the test and the lowest value at its beginning. L. monocytogenes always decreased in unripened acid-curd cheeses but showed extended growth in 21 out of 25 batches of ripened soft cheese. Contrasting results were obtained for semi-hard cheeses, as important intra- and inter-batch variability was observed. For the latter, the recommended method based on medians to calculate the growth potential led to erroneous food safety considerations, and it should always be advised to focus on absolute levels.