DNA fingerprinting allows the verification of conventional methods used to implement beef traceability. At any point along the supply chain, the identity of an animal or piece of meat can be checked by comparison of its DNA profile with an initial sample. Practical application of DNA fingerprinting to trace beef requires a choice of DNA markers as well as the optimization of sampling methods. This has been achieved as the result of collaboration between meat technicians and geneticists over a period of 4 years. The discrimination power of nine highly polymorphic microsatellite markers was evaluated. We propose that three markers (with a 0.001 probability that two individual profiles match by chance) are adequate for routine tests. Two key points along the production-commercialization chain where sampling must be systematic were defined: (i) the tagging of the calf (identity control) and (ii) after slaughter (slaughter control), before the animal loses its external appearance. The identity control was blood collected on a filter paper adapted to the ear tag; the slaughter control was the tagged ear itself. These constituted the control samples, which were archived with a code matching the individual tag number. Test samples were obtained on a random basis from live animals, carcasses, and pieces of meat at cutting halls and at the retail outlet and in cases when the verification of identity was needed. The DNA profiles of the test samples and the controls were then obtained and compared, to verify either an individual identity or the origin of a piece of meat from the stated animal.