The production of intrathecal antibodies is considered a highly specific marker for an infection of the central nervous system (CNS), e.g. borreliosis or tick-borne encephalitis (TBE). To investigate the validity of this assumption, we examined records of patients who had been hospitalized between 1989 and 1995, who were tested for borreliosis (n = 8003) and TBE (n = 904) and whose cerebrospinal fluid (CSF) had subsequently tested positive for intrathecal production of antibodies. The time period between the beginning of the symptoms and the time of the CSF examination ranged from one day to six weeks. Seventy-seven patients showed a production of intrathecal antibodies against Borrelia burgdorferi. Three of these patients were false positives with no history and no clinical signs of neuroborreliosis. In two cases, this was due to a non-specific cross-reaction caused by a preceding infection with syphilis. The third false positive was possibly caused by an earlier administration of immunoglobulins. Three patients showed a production of intrathecal antibodies against TBE virus. Two of these patients were false positives. In one case, we suspect that the production of intrathecal antibodies was caused by a non-specific immune reaction during an acute neuroborreliosis. One year earlier, the patient had contact with TBE virus through a vaccination against TBE. The cause of the second false positive is unclear, the clinical findings, acute encephalitis and the serological analysis suggest a cross-reaction with a virus similar to TBE. A specific intrathecal production of antibodies is not a proof for an infection of the CNS. In unclear cases, one should carry out a Western blot analysis or, if one suspects a case of TBE, a neutralization test.