We set out to complete the audit cycle of caesarean deliveries in order to determine if improvements could be achieved. This was a prospective review of clinical notes in a Central London Teaching Hospital. The study involved 152 women delivering between 18 May and 23 August 1998, and 226 women over the same time period in 1999. For each case, a proforma was completed within 72 hours of delivery. We recorded the total and emergency section rate; indications; decision-to-delivery times; reasons for delay; prescription of ranitidine and heparin; fetal blood samples and cord pH values. The total caesarean section rate decreased from 20.9% to 19.2%. The emergency section rate was unchanged at 14.8% in 1998 (70.9% of total) and 13.6% in 1999 (70.8% of total). Main reasons for emergency sections were failure to progress (59% in 1998, 47% in 1999) and fetal distress (27% in 1998, 34% in 1999). For failure to progress 76% of cases were performed within 1 hour in 1998 vs. 64% of cases in 1999. For fetal distress 39% of cases were delivered within 30 minutes in both years. Fetal blood samples were taken in 41% of fetal distress cases in 1998 and 34% in 1999. Cord pH was documented in 60% of emergency cases in 1998 (96% in 1999). Prescription of ranitidine rose from 53% to 81%. Heparin was well prescribed in both years (88% vs. 87%). Following the initial audit, the total caesarean section rate was significantly lower but there was no difference in the emergency section rate. The implementations had no effect on decision-to-delivery times or use of fetal blood sampling. Improvements were seen in obtaining cord pH values and ranitidine prescription.