The chief aim of our work has been to create a computer Cardiotocographic Assessment System (CAS) and thus eliminate the intra- and inter-observer variability of the visual assessment of the cardiotocogram (CTG), and to improve the assessment of the CTGs to the standard of the most experienced obstetricians. The purpose of this paper is to present the accuracy of the prediction of fetal outcome obtained by the CAS and to compare it with 4 experienced obstetricians' accuracy. Fifty CTGs from the last 30 min. of labor were assessed as normal or pathological by the computer and by the obstetricians. The condition of the newborn was evaluated from the one-minute Apgar score, the umbilical artery pH and standard base excess, and the need for resuscitation. The accuracy of the CAS was significantly better than the accuracy of the obstetrician obtaining the best result. We conclude that the development of a computer-aided cardiotocographic assessment system is worthwhile.