GSIS is often measured in the sheep fetus by a square-wave hyperglycemic clamp, but maximal β-cell responsiveness and effects of fetal number and sex difference have not been fully evaluated. We determined the dose-response curve for GSIS in fetal sheep (0.9 of gestation) by increasing plasma glucose from euglycemia in a stepwise fashion. The glucose-insulin response was best fit by curvilinear third-order polynomial equations for singletons (y = 0.018x(3) - 0.26x(2) + 1.2x - 0.64) and twins (y = -0.012x(3) + 0.043x(2) + 0.40x - 0.16). In singles, maximal insulin secretion was achieved at 3.4 ± 0.2 mmol/l glucose but began to plateau after 2.4 ± 0.2 mmol/l glucose (90% of maximum), whereas the maximum for twins was reached at 4.8 ± 0.4 mmol/l glucose. In twin (n = 18) and singleton (n = 49) fetuses, GSIS was determined with a square-wave hyperglycemic clamp >2.4 mmol/l glucose. Twins had a lower basal glucose concentration, and plasma insulin concentrations were 59 (P < 0.01) and 43% (P < 0.05) lower in twins than singletons during the euglycemic and hyperglycemic periods, respectively. The basal glucose/insulin ratio was approximately doubled in twins vs. singles (P < 0.001), indicating greater insulin sensitivity. In a separate cohort of fetuses, twins (n = 8) had lower body weight (P < 0.05) and β-cell mass (P < 0.01) than singleton fetuses (n = 7) as a result of smaller pancreata (P < 0.01) and a positive correlation (P < 0.05) between insulin immunopositive area and fetal weight (P < 0.05). No effects of sex difference on GSIS or β-cell mass were observed. These findings indicate that insulin secretion is less responsive to physiological glucose concentrations in twins, due in part to less β-cell mass.