Bovine tuberculosis (TB) is a major worldwide zoonotic disease. Foremost of the drawbacks in the control campaigns is the slow growth of its causative agent, M. bovis, as bacteriology remains the "gold standard" for the diagnosis of this disease. Rapid alternative molecular biology methods for TB diagnosis have long been hampered due to mycobacterial-linked difficulties when conventional DNA extraction techniques are applied. Moreover, the correct specificity is difficult to achieve because of the similarities in genetic background between M. bovis and other ubiquitous mycobacterial species. Nevertheless, much technological progress has been achieved in recent years, allowing the development of accurate molecular diagnosis. One of the main problems for bovine TB control is the existence of M. bovis wildlife reservoirs, a source of re-contamination in bovine TB-free herds. PCR seems an interesting alternative method for rapidly screening these species in epidemiological enquiries and immediate decision-making to avoid transmission to livestock. We describe here the validation process for a PCR diagnostic method compared to bacteriology in a wildlife TB survey.