Hydrogen gas (H(2)) has been shown to ameliorate brain injury in experimental adult rat focal ischemia and in a mild neonatal hypoxia-ischemia (HI, 90 min hypoxia) rat model. In this study we tested H(2) in moderate (120 min hypoxia) and severe (150 min hypoxia) neonatal HI rat models. We hypothesized that H(2) would improve outcomes after neonatal HI by scavenging free radicals. Two hundred (200) unsexed Sprague-Dawley rats at day 10 of life (p10) underwent neonatal HI with the Rice-Vannucci model. Multiple treatment protocols were studied, including pre-ischemic treatment, intra-ischemic treatment, and post-ischemic treatment (Sham n=32, HI n=82, HI+H(2)n=86). We also tested H(2) in middle cerebral artery occlusion (MCAO) in adult rats (MCAO n=9, MCAO+H(2)n=7) for comparison. Analysis at 24 h included infarction volume, measurement of brain concentration of malondialdehyde (MDA) (an end-product of lipid peroxidation), daily weight, Nissl histology, and mortality. In moderate and severe neonatal HI models, hydrogen gas therapy (2.9% concentration H(2)) was not associated with decreased volume of infarction or decreased concentration of MDA. H(2) gas pretreatment (2.9%) was associated with increased infarction volume in neonatal HI. In MCAO in adult rats, H(2) gas therapy demonstrated a trend of beneficial effect. Exposure of H(2) gas to non-ischemic neonates resulted in a significant increase in brain concentration of MDA. We conclude that 2.9% H(2) gas therapy does not ameliorate moderate to severe ischemic damage in neonatal hypoxia-ischemia.