OBJECTIVE: To determine the incidence of occult cerebrospinal fluid leaks (CSF) after functional endoscopic sinus surgery (FESS) and to evaluate the diagnostic performance of beta2-transferrin in blood-contaminated conditions. STUDY DESIGN: Prospective cohort study. METHODS: An analysis of 57 intraoperative samples using hydrogel 6 beta2-transferrin assay after FESS was undertaken. In case of CSF positive samples and continuing rhinorrhea, reanalysis after more than 1 year was conducted. In-vivo analysis of a primary spontaneous CSF leak sample took place to verify difficulties in detecting beta2-transferrin in blood-contaminated settings. Own titrations were performed to evaluate detection limits of CSF by beta2-transferrin and beta-trace protein assays in these settings. RESULTS: An incidence of 13% for occult CSF leaks after FESS was found. In blood-contaminated conditions, routine beta2-transferrin assays showed low sensitivity. In over 1 year follow-up, all samples were negative for CSF and none of them developed clinical relevant CSF leaks or meningitis. CONCLUSION: Occult and clinically irrelevant CSF leaks do occur in a significant proportion of patients during and shortly after FESS. Intra- and postoperatively, routine beta2-transferrin assays show low sensitivity. They should not be used in these settings. The clinical course of patients with occult CSF leaks indicated possibility of an uneventful follow-up.