The two largest Norwegian obstetrical departments, in the cities of Bergen and Stavanger, differ markedly with regard to the frequency of Cesarean section. During the 1970's the proportions rose in both departments, but the Stavanger rate remained about half of that in Bergen, the latter following the national average. These differences were not reflected in the perinatal mortality, which had the same rates and fell equally in both hospitals. To search for explanations, all records of Cesarean section cases from the years 1974 and 1979 were studied in both hospitals. We found that mechanical indications, meaning disproportion and prolonged labor, accounted for most of the difference, whereas there was no significant rise in the frequencies for the indication fetal asphyxia in either of the hospitals, these being numerically the most important indications. In fact, there was no significant rise for mechanical indications in Stavanger between 1974 and 1979, whereas in Bergen the increased frequency for this indication gave the highest significance level for any of the seven subgroups of indications. It is suggested that the chain of command may be an important cause of the observed differences between the two departments, although this is not reflected in any stated or conscious policy in either place.