For antepartum evaluation of high-risk pregnancies the non-stress test (NST) was performed in order to predict favorable or adverse fetal outcome. Between January 1, 1980 and December 31, 1980, 4078 NSTs were performed on 454 women with high-risk pregnancies. The NSTs were evaluated for pathology by a CTG pathology score. The NST interpretation were normal 95%, slight pathological 4% and severe pathological 1%. The group of women in whom all NST results were normal gave birth to healthy babies. In tact with an increasing number of pathologica-NSTs and with worsening CTG pathology score, a significant increase was found for cesarean section rate, acute operative delivery, low Apgar score, low umbilical cord artery pH and infants born small for gestational age or clinically dysmature. The frequence of perinatal morbidity was almost the same, irrespective of whether the NSTs showed accelarations or not on one or more occasions. To identify fetuses suffering from retarded intra-uterine growth the predictive value of normal NST was 97.3% and the predictive value of pathological NST was 34.8%. To identify perinatal morbidity the predictive values were 87.4% and 55.1% respectively. The NST appears to be a reliable test for antepartum assessment of fetal well-being in high-risk pregnancies.