Changes in snowpack induced by climate change can profoundly affect forest litter decomposition. A snow-depth manipulation experiment with three treatments ( i.e.,control,snowaddition, and snow-removal) was conducted to assess the effects of snow-depth changes on leaf litter decomposition of two temperate tree species [Manchurian ash ( Fraxinus mandshurica) and Dahurian larch ( Larix gmelinii) ]. The annual loss of the litter mass after one year decomposition varied between 51.3% and 57.4% for the ash and between 21.7% and 31.4% for the larch. The decomposition constants ( k) ranged from 0.048 to 0.057 and from 0.022 to 0.030 for these two species respectively. The greatest k value occurred under the snow addition treatment,while the least occurred under the snow removal treatment. Snow addition treatment shortened the 50% and 95% decomposition time by 1.1 months and 4.2 months for the ash,respectively,and by 3.7 months and 15.5 months for the larch,respectively. The snow removal treatment lengthened those decomposition time by 1.8 months and 6.4 months for the ash,and by 5.0 months and 21.1 months for the larch, respectively. Litter decomposition rate was significantly correlated with tree species,snow-depth, decomposition time,and soil temperature,but its major influencing factors varied with decomposition stage. Soil temperature and the initial litter quality were the major factors affecting decomposition rates during the snow-covered and following snow-free periods,respectively. Our findings highlight that changes in snow depth exert significantly instantaneous and prolonged effects on forest litter decomposition.