The diagnosis of tuberculous meningitis is difficult, especially when there is no extrameningial disease, because the direct examination of the cerebral spinal fluid (CSF) is often negative and the results from culture of Mycobacterium tuberculosis may take 3 to 8 weeks to become available. The morbidity and mortality remain high. So meningitis with a lymphocytic pleiocytosis and a low glucose level of CSF enjoins initiating antituberculous therapy. A pleiocytosis consisting predominantly of polymorphonuclear leucocytes in the CSF is well recognized in the early stages of the treatment, but its apearance later is hardly reported and suggests several hypotheses. We report a new case of later cytological change during antituberculous therapy.