The effect of ewe fatness on fetal weight at term in ewes underfed in late pregnancy was tested by minimising the confounding effect of differences between fatness groups in placental weight. Twin-bearing Merino ewes in a fat (n = 9, condition score 3.8 units) or moderate (n = 9, condition score 2.9) body condition were underfed to 0.6 of their requirements for energy maintenance from Day 108 to 144 of pregnancy. The fatness groups were developed over an 80-day period prior to mating by splitting a flock into 2 groups, each of similar mean liveweight and body condition score, and then enhancing the nutrient intake of one whilst restricting that of the other to maintenance levels until mating. After mating, both groups were fed similarly until Day 108 of pregnancy in an attempt to allow the development of placentas of a similar size in each. Maternal fat and protein mobilised between Day 108 and 144 of pregnancy were calculated on a per animal basis as the difference between maternal fat and protein weights at Day 108, estimated by the tritiated water dilution technique, and at Day 144, estimated by chemical analysis of fat and protein tissue following slaughter. At slaughter, the fat ewes had placentas of a size similar to those of the leaner ewes (588 v. 507 g, pooled s.e. = 50.7, P > 0.10) but produced a total weight of fetus that was 14% heavier (6646 v. 5826 g, pooled s.e. = 257.6, P <0.05). The fat ewes also mobilised more body fat between Day 108 and 144 (3.62 v. 2.58 kg, pooled s.e. = 0.446, P <0.10), and maintained higher levels of non-esterified fatty acids in their plasma during late pregnancy. There was no effect of ewe fatness on the extent of protein mobilisation (0.66 v. 0.62 kg, pooled s.e. = 0.623, P > 0.10) over late pregnancy. We conclude that additional maternal fat reserves can limit the extent to which fetal growth is restricted in ewes undernourished during late pregnancy.