Delayed cerebrospinal fluid sterilization is defined by a positive second lumbar puncture, recommanded according to the guidelines from the French Consensus Conference of 1996 between the 36th and 48th hours after the beginning of antibiotics prescribed for pneumococcal meningitis. The aim of this study was to analyze specifically delayed cerebrospinal fluid sterilization, identified during the first 5 years of the French observatory of children bacterial meningitis. The Groupe de Pathologie Infectieuse Pédiatrique (GPIP) and Association Clinique et Thérapeutique Infantile du Val de Marne (ACTIV) has set up since the first of January 2001 a descriptive national multicenter network, to determine incidence, main characteristics, and prognosis of bacterial meningitis in children. A questionnaire, available in all paediatric unit taking care of bacterial meningitis, was completed by a referral doctor. It contained reasons for inclusion in the study, anamnesis, clinical examination, treatment, pneumococcal characteristics, and short term prognosis. Delayed cerebrospinal fluid sterilization were identified, and the analysis of their medical records completed the questionnaire. From 1st January 2001 to 31 December 2005, 616 pneumococcal meningitis were identified. Among them, 442 had a second lumbar puncture, and 8 had delayed cerebrospinal fluid sterilization. The analyis of their medical records were reviewed to describe their characteristics. Two had an osteomeningeal breach, one a ventriculoperitoneal valve. All received previously an antibiotic, and were treated with a curative antibiotic by cephalosporins. Vancomycin was given in 6 cases. The antibiotic was inadapted to the French guidelines for 1 patient. There are 4 vaccine type pneumococci and only 1 strain was resistant to penicillin, and intermediate to cephalosporins. The controlled lumbar puncture was made between 36.5 and 179.4hours after beginning antibiotics. One patient has received a double dose of steroids. Three were in a coma, had convulsions, and were ventilated, none died. One patient has a sequellar paired deafness, two a severe disability, four a normal psychomotor development. The delay of sterilization is a rare situation and represented only 1.8 % of pneumococcal meningitis during the first five years of the observatory. These results suggest that a second lumbar puncture to assess sterilization could be proposed only in cases of unfavourable clinical course, MIC greater than or equal 0.5mg/l to 3GC, risk factors for delayed cerebrospinal fluid sterilization and high bacterial inoculum.