A prospective 6-month study in Yaounde evaluated 49 children aged from 2 months to 8 years, hospitalized with bacterial meningitis. They were randomly assigned to one of two initial treatment groups, either an ampicillin-chloramphenicol combination (group A) or chloramphenicol alone (group B). The majority of patients were infected with Haemophilus influenzae, and the majority of deaths were caused by Streptococcus pneumoniae. Altogether, 17.9% of Haemophilus influenzae isolates were ampicillin-resistant and 3.6% chloramphenicol-resistant. We found no isolate resistant to both antibiotics. Response to both treatments was similar in both groups. The theoretical risk of treatment failure with ampicillin was higher than with the ampicillin-chloramphenicol combination (p less than 0.05). There was no statistically significant difference between the risk of treatment failure with the ampicillin-chloramphenicol combination and the risk with chloramphenicol alone (p less than 0.05), but the latter was increased by the occurrence of chloramphenicol-resistant isolates of Streptococcus pneumoniae (11.1%). Although treatment with an ampicillin-chloramphenicol combination is four times more expensive than treatment with chloramphenicol alone, costwise it is also one-quarter the price of a third-generation cephalosporin (moxalactam). At present, the ampicillin-chloramphenicol combination can be suggested as the first choice for initial treatment considering both the epidemiological data and the cost/efficiency ratio in the area of Yaounde.