Previous studies in near-term sheep have shown that maternal cocaine injection causes acute fetal cerebral vasodilation along with transient hypoxemia and hypertension. Preterm sheep fetuses have lower cerebral O2 consumption (CMRO2) and their cerebrovascular responses to hypoxemia are attenuated compared with near-term fetuses. We therefore tested the hypothesis that fetal cerebrovascular responses to maternal cocaine injection may also differ earlier in gestation. We studied nine immature fetal sheep at 0.65 gestation using the same experimental protocol we used in previous studies in near-term sheep. Fetal studies were done in utero, 2 d after vascular catheter placement. We measured cerebral blood flow (CBF) using microspheres, arterial and sagittal sinus O2 content, and cocaine concentrations. We calculated cerebrovascular resistance (CVR) as mean arterial blood pressure divided by CBF. Measurements were made before and 2, 5, and 15 min after a 2 mg/kg maternal cocaine injection. At 2 min, fetal Cao2 decreased (18 +/- 6%, mean +/- SEM), and there was cerebral vasoconstriction (CVR increased by 22 +/- 5%). At 5 min, CBF increased (19 +/- 9%), but because blood pressure increased also, CVR returned to baseline, and therefore there was no vasodilation compared with baseline. Furthermore, at 5 min there was a 22 +/- 6% decrease in Cao2 and a 21 +/- 6% increase in mean arterial blood pressure. There were no changes in CMRo2 throughout the study, but at 2 min, cerebral O2 delivery decreased. Differences in cerebrovascular responses to maternal cocaine injection earlier in gestation may be due to differences in vascular development and/or to developmental differences in responses to cocaine, cocaine metabolites, and/or to hypoxemia.