Patients' choice may be an obvious argument to defend the high rate of caesarean sections in Brazil. Potter and colleagues (2001) show that we may not have this easy escape hatch available to us. They also present data indicating that multiparous women who delivered within the public sector and were certain that they did not want a caesarean section were not at particular high risk (8.5%) of undergoing one (this was provided that they had no previous caesarean section). This suggests that women who know from their own experience that they can deliver vaginally have a low caesarean section rate. The causes of the high caesarean section rates in the public sector may be less related to what obstetricians do than to what they don't do: encouraging women to believe in their ability to have a vaginal delivery.