We investigated the effect of restoration of normoglycaemia or normoinsulinaemia in fetuses of starved ewes on plasma IGF-I and IGF-II concentrations. Paired maternal and fetal blood samples were taken during an initial 2-day control period, after 48 h of maternal starvation, during 24 h fetal infusion of glucose (n = 6) or insulin (n = 4) while maintaining maternal starvation and after 48 h maternal refeeding. After 48 h starvation maternal and fetal plasma IGF-I, insulin and blood glucose fell (maternal IGF-I 38.9 +/- 3.6 to 16.4 +/- 1.8 nM and fetal IGF-I 13.2 +/- 0.8 to 7.1 +/- 0.7 nM, both P < 0.05). Fetal plasma IGF-II also fell (147.8 +/- 9.1 to 112.2 +/- 3.8 nM, P < 0.05), but maternal plasma IGF-II rose (71.8 +/- 6.3 to 88.8 +/- 9.2 nM, P = 0.10). Fetal glucose replacement raised fetal plasma IGF-I (11.4 +/- 1.2 nM), IGF-II (149.7 +/- 6.5 nM), insulin and blood glucose to near control values (all P < 0.05). Fetal insulin replacement raised fetal plasma IGF-I (9.0 +/- 0.6 nM) and insulin (all P < 0.05) while IGF-II (105.2 +/- 8.4 nM) and blood glucose remained depressed. Neither fetal infusion had any significant effect on maternal plasma IGF-I (13.1 +/- 1.6 nM), IGF-II (77.5 +/- 8.7 nM), insulin or blood glucose. After 48 h maternal refeeding fetal IGF-I (12.4 +/- 0.4 nM), fetal IGF-II (158.4 +/- 8.9 nM), maternal IGF-II (67.1 +/- 3.0 nM), maternal and fetal insulin and glucose had returned to near control values in both groups. Maternal IGF-I remained below control values (24.7 +/- 2.5 nM, P < 0.05). The data suggest that fetal IGF-I and IGF-II are independently regulated in the fetal circulation. While glucose plays an important role in the regulation of both IGF-I and IGF-II, the influence of glucose on fetal IGF-I is likely to be mediated by insulin, whereas for IGF-II the effect of glucose is insulin-independent.