A three month prospective audit of wound infection following emergency and elective caesarean section was carried out in five West Yorkshire hospitals. Among 4076 women undergoing delivery in the five obstetric departments, the caesarean rate was 15.4%. The overall infection rate was 45/628 (7.2%) with a range of 2.5-17.2% between the five centres. The infection rate was 14/226 (6.2%) when antibiotics were used compared with 31/402 (7.7%) without antibiotics. The use of prophylactic antibiotics made no significant difference to the infection rate, which did not correlate with duration of labour or of ruptured membranes. The number of vaginal examinations correlated with the infection rate. In conclusion, the caesarean section rate observed was higher than that estimated for the UK as a whole, but was distorted by one centre with a high rate. For the other four hospitals the caesarean rate was unexceptional. The ratio of emergency to elective operations was comparable with recently reported values in the UK and the wound infection rate was within the widely varying limits found in previous studies. In view of the relatively low infection rate recorded without antibiotics, in the interests of cost effectiveness, prophylaxis may be limited in future to selected women at high risk. Because this was an audit rather than a randomized study we cannot exclude that this is already happening on an empirical basis.