Lyme serology is too commonly prescribed in absence of a well established differential diagnosis. Its low positive predictive value is very often unusable. IgG and IgM positive results are tricky to interpret and often lead to clinically unjustified antibiotherapy. Clinical and epidemiological approaches must integrate the serological results. The microbiologist should have all available information to be able to communicate an appropriate interpretation with the results. This way can propose adequate analysis leading to higher predictive values. We present essentials and pitfalls of the Lyme serology, suggesting new possibilities and reminding physicians as to which tests have not yet brought any proof of benefit to the patient.