The effects on the fetus and placenta of graded reductions of uterine blood flow to 30-90% of control have been studied in sheep at days 125-143 of pregnancy. Reduction of uterine flow to 70-90% of control had little effect upon fetal oxygenation or heart rate or blood pressure but elevated fetal plasma catecholamine concentration. Reduction of flow to 30-50% of control depressed fetal arterial and umbilical venous PO2 but had little effect upon oxygen consumption unless the umbilical venous value fell below about 14 mmHg when it was depressed by up to 30%. Placental oxygen consumption did not fall and was therefore maintained at the expense of the fetus. Fetal arterial pressure rose by 10-12 mmHg and heart rate fell by about 30 beats/min during the first 10-15 min then rose above its initial value. Plasma adrenaline and noradrenaline concentrations rose progressively at a rate which increased with greater degrees of asphyxia. When uterine blood flow was reduced below one-half of normal, net placental consumption of glucose fell and there was evidence of substantial provision of glucose and lactate from the fetus. Fetal production of lactate increased sharply and much of this appeared to be consumed by the placenta at a rate sufficient to account entirely for the deficit in net glucose consumption. The results are consistent with the hypothesis that the fetus senses even small changes in uterine blood flow that are alone insufficient to elicit significant blood gas changes. When the fall in uterine flow caused by arterial compression is relatively large, nutrient supply to the placenta is maintained at the expense of the fetus and as a result of fetal glucose and lactate production. The elevation of fetal arterial PCO2 appears to enhance fetal responses to hypoxia. The results are discussed in relation to the fetal responses to brief and prolonged reductions in uterine blood flow.