Abstract: Objective To investigate the effects of hydrogen on the lung damage of mice at early stage of severe burn. Methods One hundred and sixty ICR mice were divided into sham injury, hydrogen, pure burn, and burn+ hydrogen groups according to the random number table, with 40 mice in each group. Mice in pure burn group and burn+ hydrogen group were inflicted with 40% total body surface area full-thickness scald (hereafter referred to as burn) on the back, while mice in sham injury group and hydrogen group were sham injured. Mice in hydrogen group and burn+ hydrogen group inhaled 2% hydrogen for 1 h at post injury hour (PIH) 1 and 6, respectively, while mice in sham injury group and pure burn group inhaled air for 1 h. At PIH 24, lung tissue of six mice in each group was harvested, and then pathological changes of lung tissue were observed by HE staining and the lung tissue injury pathological score was calculated. Inferior vena cava blood and lung tissue of other eight mice in each group were obtained, and then content of high mobility group box 1 (HMGB1) and interleukin-6 (IL-6) in serum and lung tissue was determined by enzyme-linked immunosorbent assay. Activity of superoxide dismutase (SOD) in serum and lung tissue was detected by spectrophotometry. After arterial blood of other six mice in each group was collected for detection of arterial partial pressure of oxygen (PaO₂), the wet and dry weight of lung tissue were weighted to calculate lung wet to dry weight ratio. The survival rates of the other twenty mice in each group during post injury days 7 were calculated. Data were processed with one-way analysis of variance, LSD test and log-rank test. Results (1) At PIH 24, lung tissue of mice in sham injury group and hydrogen group showed no abnormality. Mice in pure burn group were with pulmonary interstitial edema, serious rupture of alveolar capillary wall, and infiltration of a large number of inflammatory cells. Mice in burn+ hydrogen group were with mild pulmonary interstitial edema, alveolar capillary congestion accompanied by slight rupture and bleeding, and the number of infiltration of inflammatory cells was smaller than that in pure burn group. The lung tissue injury pathological scores of mice in sham injury group, hydrogen group, pure burn group, and burn+ hydrogen group were (0.7±0.5), (0.8±0.5), (6.1±1.0), and (2.8±0.8) points, respectively. The lung tissue injury pathological score of mice in pure burn group was significantly higher than that in sham injury group (P<0.001). The lung tissue injury pathological score of mice in burn+ hydrogen group was significantly lower than that in pure burn group (P<0.001). (2) At PIH 24, the content of HMGB1 and IL-6 in serum and lung tissue of mice in hydrogen group was close to that in sham injury group (with P values above 0.05). The content of HMGB1 and IL-6 in serum and lung tissue of mice in pure burn group was significantly higher than that in sham injury group (with P values below 0.001). The content of HMGB1 and IL-6 in serum and lung tissue of mice in burn+ hydrogen group was significantly lower than that in pure burn group (with P values below 0.001). (3) At PIH 24, the activity of SOD in serum and lung tissue of mice in hydrogen group was close to that in sham injury group (with P values above 0.05). The activity of SOD in serum and lung tissue of mice in pure burn group was significantly lower than that in sham injury group (with P values below 0.001). The activity of SOD in serum and lung tissue of mice in burn+ hydrogen group was significantly higher than that in pure burn group (with P values below 0.001). (4) At PIH 24, there was no statistically significant difference in PaO₂ among the mice in four groups (F=0.04, P>0.05). (5) At PIH 24, the ratios of lung wet to dry weight of mice in sham injury, hydrogen, pure burn, and burn+ hydrogen groups were 3.52±0.22, 3.61±0.24, 7.24±0.32, and 5.21±0.41, respectively. The ratio of lung wet to dry weight of mice in pure burn group was significantly higher than that in sham injury group (P<0.001). The ratio of lung wet to dry weight of mice in burn+ hydrogen group was significantly lower than that in pure burn group (P<0.001). (6) The survival rates of mice in sham injury group and hydrogen group during post injury days 7 were 100%. Compared with those in sham injury group, survival rates of mice in pure burn group from post injury days 3 to 7 were significantly decreased (with P values below 0.05). Compared with those in pure burn group, survival rates of mice in burn+ hydrogen group from post injury days 5 to 7 were significantly increased (with P values below 0.05). Conclusions Hydrogen can significantly alleviate the infiltration of inflammatory cells and improve the pathological lesions of lung tissue of mice with severe burn. It has the effects of reducing inflammatory reaction and inhibiting oxidative stress, further showing the protective effect on the lung of burn mice.