A previously described polymerase chain reaction (PCR) assay (B. Furrer, U. Candrian, C. Höfelein, and J. Lüthy, J. Appl. Bacteriol. 70:372-379, 1991) was used to analyze food for the presence of Listeria monocytogenes. Food samples were artificially contaminated to develop two procedures to detect the organism following enrichment steps. Procedure A was based on dilution of the enrichment broth followed by lysis of the bacteria and direct analysis of the lysate with PCR. With procedure A and artificially contaminated food samples, it was possible to detect fewer than 10 bacteria per 10 g of food. In procedure B, centrifugation was used to concentrate bacteria before lysis and PCR. With procedure A, 330 naturally contaminated food samples of several types were analyzed. Twenty samples were found to be positive for L. monocytogenes, which was in agreement with the classical culture technique. By using procedure B on a subset of 100 food samples, 14 were found to be positive by PCR whereas the classical culture method detected only 13. Analysis times, including enrichment steps, were 56 and 32 h with procedures A and B, respectively.