To assess the value of adjunctive intravenous dexamethasone (DXM) and oral glycerol (GLY) for the treatment of bacteriologically proved bacterial meningitis, 122 infants and children with bacterial meningitis were randomly assigned to receive DXM intravenously (n = 32), GLY orally (n = 30), DXM plus GLY (n = 34) or neither (n = 26) of these drugs. All patients were treated with the same antimicrobial agent, ceftriaxone. The patients were followed neurologically for as long as 6 months. A thorough hearing evaluation was performed routinely 2 months or more after discharge from hospital. Overall 4 (7%) of the GLY-treated patients, compared with 11 (19%) of those not given GLY, developed audiologic or neurologic sequelae (P = 0.052), the relative risk of sequelae being 2.94 (95% confidence interval, 0.99 to 8.72). The patients who had received GLY showed less severe or profound bilateral hearing impairment than those not given GLY (0 vs. 7%, P = 0.049), and none of them had other neurologic abnormalities 3 or 6 months after discharge, compared with 5 (9%) of those not treated with GLY (P = 0.024). The DXM recipients showed only a tendency to less severe hearing impairment than those not given DXM. In conclusion oral GLY prevented neurologic sequelae in infants and children with bacterial meningitis more effectively than intravenous DXM.