Neonatal rats were exposed to 6.6 g/kg of alcohol each day between postnatal days 4 and 10 while artificial-rearing procedures were used, in a manner which produced high peak and low trough blood alcohol concentrations each day. Gastrostomy controls were reared artificially with maltose/dextrin isocalorically substituted for alcohol in the milk formula, and suckle controls were reared normally by dams. The pups were sacrificed on day 10 and tissue sections (2 microns thick) were obtained in the sagittal plane through the cerebellum and in the horizontal plane through the hippocampal formation. Overall area measures were obtained for the hippocampus proper, area dentata, and cerebellum, along with areas of the cell layers of these regions. In the hippocampal formation, cell counts were made of the pyramidal cells of the hippocampus proper, the multiple cell types of the hilus, and the granule cells of the area dentata. In the cerebellum, cell counts of Purkinje cells, granule cells of the granular layer, granule cells of the external granular layer, and mitotic cells of the external granular layer were obtained from lobules I, V, VII, VIII, and IX. Alcohol selectively reduced areas and neuronal numbers in the cerebellum but had no significant effects on neuronal numbers in the hippocampal formation. Purkinje cells exhibited the greatest percent reductions, and cerebellar granule cells were significantly reduced in the granular layer but not in the external granular layer. All lobules showed these effects, but lobule I was significantly more affected than the other four lobules that were analyzed. The results demonstrate the differential vulnerability of selected neuronal populations to the developmental toxicity of alcohol exposure during the brain growth spurt.