OBJECTIVE: To determine whether prophylactic antibiotic administration using cefoxitin at the time of elective caesarean section significantly reduces infectious morbidity. SETTING: A tertiary teaching hospital in a large urban city in South Africa. PARTICIPANTS: Women undergoing elective caesarean section. DESIGN: A prospective, double-blind randomised placebo-controlled trial. METHODS: Four hundred and eighty women undergoing elective caesarean section had cefoxitin or placebo administration after umbilical cord clamping. Postpartum complications including febrile morbidity, wound infection, endometritis, urinary tract infection, pneumonia and transient postpartum fever were recorded, as were the duration of hospital stay and the need for therapeutic antibiotics. RESULTS: Wound infection was the most common complication occurring in 13.3% and 12.5% of women in the placebo and cefoxitin groups, respectively. Prophylactic antibiotics did not decrease febrile morbidity, wound infection, endometritis, urinary tract infection and pneumonia. Women who received cefoxitin stayed on average a day less in hospital than those who received placebo (6.9 vs 7.8 days, risk difference 0.94 CI 1.57 - 0.31 days). Eleven women (4.6%) in the placebo group and eight (3.4%) in the cefoxitin group had microbiological evidence of wound infection. Staphylococcus aureus was the most common pathogen (43%) isolated. Similar proportions in both groups (6.3% placebo and 5.1% cefoxitin) required a course of therapeutic antibiotics. CONCLUSION: Antibiotic prophylaxis with cefoxitin in elective caesarean section did not reduce post-operative infectious morbidity in this double-blind randomised placebo controlled trial.