turition, little is known of the source of this oestrogen rise in the cow. It seemed pertinent, therefore, to determine the levels of oestrogens in fetal and maternal plasma over the last 3 weeks of gestation. In the present paper, the levels of free oestrone and oestradiol-17\g=b\ in the maternal jugular and utero-ovarian vein plasma and in fetal vena cava plasma in relation to the initiation of parturition in the cow were determined by radioimmunoassay.Catheters were implanted surgically into the maternal utero-ovarian and jugular veins and into the fetal posterior vena cava through the fetal saphenous vein in three cows during the last third of gestation. Fetal and maternal blood samples were taken twice daily until parturition; when calving was imminent, maternal utero-ovarian samples were taken more frequently. Blood was collected into ice-cold, heparinized syringes, centrifuged immediately and the plasma was stored at -15\s=deg\C until analysis. The plasma was extracted with 10 vols redistilled diethyl ether and chromatographed on micro-columns of Sephadex (Wu & Lundy, 1971). The oestrone and oestradiol fractions thus separated were subjected to radioimmunoassay by the method of Peterson & Common (1972), using antisera against oestradiol-17ß-17-BSA and oestradiol-17/J-6-rBSA, respectively. Oestrone cross-reacted 56% with the oestradiol-17ß-17-BSA antiserum and oestradiol-17a cross-reacted 1-9% with the antiserum against oestradiol-17/Í-6-BSA. Over the range of 0-01 to 1-0 ng, recoveries of oestrone and oestradiol-17 ß were of the order of 70 to 80% and the within-assay coefficient of variation was 10 to 12%. The between-assay coefficient of varia¬ tion, calculated from control samples run in duplicate with each assay was 15%. The sensitivity of the method, determined from the variability of the reagent blanks, was 5 pg/ml for oestrone and 4 pg/ml for oestradiol-17ß. The results in this paper have been corrected for procedural losses.The mean values for three cows are presented in Text-fig.