Résumé : The diagnosis of acute neuroborreliosis may be difficult if it's regarded as a "classical" infectious disease. Through a clinical case, we illustrate the difficulties met and we suggest two ways of reflexion to assist in the diagnosis:--firstly, we explain how the comprehension of the behavior of the pathogen agent, which is similar to a parasitic behavior, can help to choose and interpret the results of additional tests;--secondly, we develop practically the clinical form of Bayes's theorem to demonstrate the interest of a rational Bayesian and abductive approach which should be preferred to the classical hypothetical and deductive reasoning.