The decision to regionalize obstetric services in the Montreal metropolitan area was designed to improve the efficiency of the system. Changes in types of birth between 1981 and 1984 in this area were studied in order to evaluate the effects of this decision on cesarean section rates. Results show a sudden increase in cesarean rates, coinciding with the regionalization decision. Hypotheses put forward are linked to the characteristics of hospitals remaining after this measure, as well as the organisational disturbances which it may have caused.