One hundred physiological pregnant women and 76 pathological pregnant women suffering from gestational diabetes and pregnancy-induced hypertension underwent a cardiotocographic examination during the course of routine diagnostic tests. The interpretation of cardiotocographic printouts was carried out using traditional as well as computerised methods. The outcome of these tests was then related to neonatal outcome and other parameters which contributed to defining the prognosis of pregnancy. Computerised analysis was found to provide a more reliable diagnosis in comparison to traditional methods in identifying those pregnancies with a pathological neonatal outcome. In particular, in the group of physiological pregnancies, computerised interpretation proved more reliable in 87.5% of cases in which neonatal outcome was pathological; on the contrary, traditional interpretations only revealed 37.5% of the same cases. In pathological pregnancies, automatic interpretation was also found to be more reliable in predicting the non-pathological outcomes, whereas traditional methods provided a high incidence of uncertain answers.