There is considerable concern over the widespread use of caffeine during and after pregnancy. We have therefore examined the effect of perinatal caffeine use on the vulnerability of the immature brain to hypoxic ischemia (HI). Rat pups were exposed to caffeine during the first 7 d after birth by addition of a low or a high dose (0.3 or 0.8 g/L) of caffeine to the drinking water of their dams. At 7 d the pups were exposed to unilateral carotid occlusion + exposure to 7.70% oxygen for 100 min. The extent of HI brain damage was evaluated 2 wk after the insult. The effects of caffeine on A1and A2areceptors, A1mRNA and A2amRNA, were examined by receptor autoradiography andin situhybridization. Caffeine, theobromine, theophylline, and paraxanthine were analyzed in plasma of separate animals. Exposure to caffeine reduced HI brain damage from 40.3 ± 3.2% in controls to 29.8 ± 4.0% (p< 0.05) in low dose and 33.7 ± 3.9% (NS) in the high dose group. The A1receptor density measured as [3H]-l,3-dipropyl-8-cyclopentyl xanthine ([3H]-DPCPX) binding was not significantly affected after low dose caffeine but increased in the brain of rat pups in the high dose group. The A2areceptor density measured as [3H]-2[p-(2-carbonylethyl)-phenethylamino]-5′-N-ethylcarboxamidoadenosine ([3H]-CGS 21680) binding and the expression of A1mRNA and A2amRNA were not altered by caffeine treatment. In conclusion, low dose caffeine exposure (plasma levels corresponding to umbilical cord plasma in newborns of coffee-consuming mothers) reduced HI brain damage by 30% in 7-d-old rats. This ameliorating effect could not be accounted for by up-regulation of adenosine receptors.