Various aspects of physician bias in the decision to perform a cesarean section were investigated by comparing indications for cesarean sections for 1979 and 1980 and investigating the distribution of indications according to the day of the week. The contention that obstetricians may perform cesarean sections on weekends for secondary gain was refuted by showing no difference in the distribution of nonelective cesarean section indications between weekdays and weekends. Statistically significant changes in cesarean section indications were, however, observed between the two investigated years, which may be attributed to an organizational change within the department toward more sophisticated obstetric care. These changes included a significant increase in cesarean sections for dystocia and breech presentation and a decrease for fetal distress and "other" indications. Because the total cesarean section rate did not change during the study periods, it is suggested that modern obstetric practice per se does not increase cesarean section rates, but may shift the distribution of indications for cesarean section.