We have previously shown that maternal Fe deficiency not only reduces fetal size, but also increases blood pressure in the offspring when they are adults. In this paper we examine whether there are critical periods when supplementation reverses or fails to reverse the effect both on size and on expression of genes of Fe metabolism. We made dams Fe deficient, mated them and provided supplements of Fe in the diet from the beginning of gestation (0.5D), from 7.5D or from 14.5D. Within 12 hrs of birth, dams and neonates were killed and tissues taken and examined. Fe deficiency throughout pregnancy reduces neonatal size. Supplementation from the beginning of the first, second or third week all reduced the effect. Maternal haematocrit was restored to normal levels only in animals given supplements for at least two weeks. In contrast, the neonates' Fe levels were normal in all supplemented groups. These results were mirrored in liver Fe levels and in transferrin receptor mRNA. IRE-regulated DMT1 increased in maternal and neonatal liver. Non-IRE regulated DMT1 levels did not change in the maternal liver, but decreased in the neonatal liver. H and L ferritin mRNA levels also showed different patterns in the mother and her offspring. Finally, the neonatal size correlated with maternal Fe stores, and not with those of the fetus. The data demonstrate that Fe supplementation during pregnancy is most effective when given in the first trimester, rather than later in gestation.