On a population-based level, the incidence of reactive arthritis (ReA) is 0.6-27/100,000. The definition of ReA varies and its pathogenesis is not yet clear. Attempts in basic immunology to suggest hypotheses for proliferation of forbidden B cell clones, molecular mimicry and involvement of cross-reactive antibodies are helpful but not sufficient. Importantly, for the clinical diagnosis of the preceding infection, serology is widely used. Unfortunately, the accuracy of associations between serologic findings and clinical conclusions is plagued by poor standardisation of methods. So far, few attempts have been done to examine the pitfalls of different approaches. Here, we review several serologic techniques, their performance and limitations. We will focus only on serology for Yersinia, Campylobacter, Salmonella and Chlamydia trachomatis because these bacteria have a longer history of being associated with ReA. We also address controversies regarding the role of serology for some other bacteria linked to autoimmune disorders.