This study was a systematic anonymous audit of routinely collected data in a tertiary referral obstetric unit in London and included data from deliveries over a 10-year period (1992-2001). Data for all caesarean sections at full dilatation were collected, including maternal demographic information, the grade of operating clinician, and the place of delivery. Neonatal data collected included birth weight and umbilical arterial pH. No changes in the demographics of the population were observed. No increased rates of malposition were observed. Birth weight did not change. Increasing preference for the ventouse over forceps (ratio 0.2:1 to 1.9:1) over the decade (p = 0.002) was seen with an increased tendency to conduct the delivery in the operating theatre (p = 0.0025). Rate of caesarean section at full dilatation increased (2% by 2001). Increasing failures of operative vaginal delivery, especially using the ventouse (regression coefficient p = 0.025), and reduced attempts at instrumentation (regression coefficient p = 0.002) were seen.