A previously published rapid (<5 h) proof-of-concept protocol for the concentration of the foodborne pathogen Listeria monocytogenes from milk, based on the solubilization of the food matrix, was further evaluated. The original protocol was modified to detect gram-negative and other gram-positive bacteria and to broaden the range of food matrices by using Lutensol instead of sodium dodecyl sulfate as the main detergent in the buffer. A new protocol using a protease and sucrose buffer was established for the analysis of meat and fish. Matrix lysis was used for dairy products, ice cream, milk, fish, meat, eggs, and blood. Solubilization of the foodstuffs resulted in bacterial pellets of reasonable size for further quantification. Using L. monocytogenes, Staphylococccus aureus, Bacillus cereus, Escherichia coli, and Salmonella Typhimurium as model organisms, microscopic analysis of the remaining bacterial pellets revealed that the recovered bacteria remained physically intact, albeit their viability was compromised. In model experiments using free DNA, the free target DNA was reduced by 5 log units. The compatibility of matrix lysis with subsequent real-time PCR analysis has been demonstrated with salmon, chicken, egg, ice cream, cheese, and blood samples that were artificially contaminated with L. monocytogenes, S. aureus, and Salmonella Typhimurium. These experiments resulted in an average recovery of 48.7% (relative standard error, 83.4%) of the inoculated bacterial cells with the real-time PCR assay. The average detection limit of the method was 7.3 CFU/ml for all examined foodstuffs and bacterial target organisms.