In the recently published Peristat study, in which perinatal mortality in the countries of the European Union is compared, the figures for The Netherlands are higher than in the other countries. These figures are based partly on civil registration data. In the countries where these civil registration data were compared with clinical records, i.e. The Netherlands, Northern Ireland and Greece, a considerable underregistration was demonstrated. In The Netherlands, prenatal screening for congenital anomalies is less prevalent than in some other European countries and paediatricians are more reluctant to resuscitate seriously immature infants. This may have some influence on perinatal mortality figures, but a more fundamental question is: are the figures from various countries, based on national civil registration and a large number of divergent clinical registration systems, comparable? This question has not been answered satisfactorily.