T his study was designed to determine the potential of IGF-1 as a neuronal rescue agent after cerebral ischemia. Unanesthetized late gestation fetal sheep were subjected to 30-min cerebral ischemia by inflation of carotid artery occluder cuffs. 2 h later either 0.1 microgram rhIGF-1, 1 microgram rhIGF-1, 10 micrograms rhIGF-1, or vehicle was infused into a lateral cerebral ventricle over 1 h. Histologic outcome was assessed 5 d later. Overall neuronal loss was reduced with 0.1 microgram (P < 0.05) and 1 microgram (P < 0.002) rhIGF-1, but treatment with 10 micrograms was not effective. With 1 microgram rhIGF-1 neuronal loss scores were significantly lower in brain regions examined including cortex, hippocampus, and striatum, whereas with 0.1 microgram rhIGF-1 the parietal cortex and thalamus were not improved and the improvement seen in other regions was less than with 1 microgram rhIGF-1. Treatment with 1 microgram rhIGF-1 also delayed the onset of seizures and reduced their incidence. Moreover, the secondary phase of cytotoxic edema was reduced and delayed in onset. We conclude that low dose rhIGF-1 therapy promotes neuronal rescue after cerebral hypoxic-ischemic injury in utero, but the effect is dose dependent. Importantly, rhIGF-1 is effective and nontoxic when administered 2 h after the hypoxic ischemic insult. This distinguishes IGF-1 from most other neuroprotective therapies and suggests clinical application may be possible.