Listeria monocytogenes blood agar (LMBA) was compared to Listeria selective agar based on lithium chloride and ceftazidime (LA) and to the Oxford and Palcam media recommended by ISO and IDF for the detection and enumeration of L. monocytogenes from foodstuffs and food-processing environments. LMBA is based on trypticase soy agar with the following additions: sheep blood (5%) and as selective agents lithium chloride (10 g/l), polymyxin B sulphate (10 mg/l) and ceftazidime (20 mg/l), whereas the selectivity of LA is based on lithium chloride (15 g/l) and ceftazidime (15 g/l). The media were compared in the detection of L. monocytogenes after enrichment from naturally contaminated foodstuffs (n = 423) and from food-processing environments (n = 93), and in the enumeration of the species from naturally contaminated foodstuffs (n = 287). LMBA was superior both to the standard media and to LA in detection after enrichment and also in enumeration, except in the case of fresh broiler cut samples. The overall sensitivities of the Palcam, Oxford, LA and LMBA media were 68%, 67%, 74% and 96% in detection after enrichment and 64%, 73%, 76% and 80% in the enumeration of the species from ready to eat foods. The superiority of LMBA is based on distinguishing L. monocytogenes from other Listeria species by detection of beta-hemolysis, whereas the other media gave false-negative results because of the overgrowth of Listeria spp. other than L. monocytogenes, especially in detection after enrichment. A more selective medium than LMBA would have been required for the enumeration of the species from samples with high levels of competitive bacteria other than Listeria spp. The results indicate the need for a more specific isolation medium for L. monocytogenes in addition to those recommended by ISO and IDF for both detection and enumeration. LMBA offers an alternative to be used in combination with either Palcam or Oxford as well as with LA.