The aim of the study was to identify risk factors associated with increased risks of post Caesarean febrile morbidity from a retrospective epidemiological analysis, and to prospectively evaluate a protocol of selective antibiotic prophylaxis based on these risk factors. Caesarean section births over a three-year period--during which no antibiotic prophylaxis protocol was practised--were reviewed. Risk factors for febrile morbidity were identified using both univariate and multivariate analysis. A protocol of selective antibiotic prophylaxis, incorporating the significant risk factors identified, was then evaluated prospectively over a nine-month period. The incidence of febrile morbidity from the retrospective study was 16.9%. A logistic regression model, with febrile morbidity as the dependent variable, identified only the duration of labour over 10 hours as a significant factor. A protocol of antibiotic prophylaxis using intravenous ampicillin was then practised in the presence of these two factors. Evaluation of this protocol after nine months showed that the incidence of febrile morbidity was reduced significantly from 20% in those not given antibiotics to around 9.6% in those eligible for and given antibiotics. However, the overall incidence of febrile morbidity remained at around 16-18%.