Abstract. The aim of the present study was to compare the usefulness of cystine aminopeptidase (CAP) and human chorionic somato‐mammotrophin (HCS) estimations in maternal serum and maternal urinary oestrogen excretion in predicting intra‐uterine growth retardation. The material consists of 43 patients who subsequently gave birth to infants with a birth weight less than 2 S.D. from the mean according to the gestational age. The patients were controlled with simultaneous analyses of CAP, HCS and urinary oestrogen assays every week from the 36th week of pregnancy. The patients have been compared with another group in which infants with normal birth weight and without signs of fetal distress were delivered. Intrauterine growth retardation (IGR) was predicted by low oestrogen levels in 58%, low HCS levels in 42% and low CAP levels in 35%. The difference between the biochemical tests was not statistically significant. IGR infants who developed fetal distress were predicted by oestrogen assays in 72%, by HCS assays in 62% and by CAP assays in 48%. If urinary oestrogen assays were combined with HCS assays this combination predicted IGR significantly better than oestrogen assays alone. The combination HCS‐CAP was as informative in this respect as urinary oestrogen assays alone. The results indicate that all methods tested are rather insensitive in predicting the total number of IGR infants. The capacity for predicting IGR infants with fetal distress is, however, rather high. For this purpose a combination of tests is preferable as no test fulfils the qualifications of being superior to the other. The combination oestrogens–HCS seems to be most suitable in this respect.