In bacterial meningitis, severe systemic and local inflammation causes long-term impairment and death of affected patients. The current antibiotic therapy relies on cell wall-active beta-lactam antibiotics, which rapidly sterilize the cerebrospinal fluid (CSF). However, beta-lactams inhibit cell wall synthesis, induce bacteriolysis, and thereby evoke a sudden release of high amounts of toxic and proinflammatory bacterial products. Because tissue damage in bacterial meningitis is the result of bacterial toxins and the inflammatory host response, any reduction of free bacterial compounds promises to prevent neuronal damage. In vitro experiments and randomized prospective animal study. University research laboratories. Streptococcus pneumoniae broth cultures and New Zealand White rabbits. We evaluated a concept to improve bacterial meningitis therapy in which a short-term pretreatment with the protein synthesis-inhibiting antibiotic rifampicin precedes the standard antibiotic therapy with ceftriaxone. First, logarithmically growing pneumococcal cultures were subdivided and exposed to different antibiotics. Then, rabbits suffering from pneumococcal meningitis were randomized to receive rifampicin pretreatment or ceftriaxone alone. In pneumococcal cultures, quantitative immunoblotting and real-time polymerase chain reaction revealed a reduced release of pneumolysin and bacterial DNA by rifampicin pretreatment for 30 minutes in comparison with ceftriaxone treatment alone. In vivo, a 1-hour rifampicin pretreatment reduced the release of bacterial products and attenuated the inflammatory host response, as demonstrated by decreased CSF levels of prostaglandin E2 and total protein and increased glucose CSF/plasma ratios. Rifampicin pretreatment reduced infection-associated neuronal apoptotic cell loss compared with ceftriaxone-treated controls. A short-term pretreatment with rifampicin reduced the beta-lactam-induced release of deleterious bacterial products, attenuated inflammation, and thereby decreased neuronal cell loss in experimental bacterial meningitis. This concept has the potential to reduce inflammation-associated neuronal injury in bacterial meningitis and should be evaluated in a clinical trial.