We have implemented a long-term snow-fence experiment at the Niwot Ridge Long-Term Ecological Research (NWT) site in the Colorado Front Range of the Rocky Mountains, U.S.A., to assess the effects of climate change on alpine ecology and biogeochemical cycles. The responses of carbon (C) and nitrogen (N) dynamics in high-elevation mountains to changes in climate are investigated by manipulating the length and duration of snow cover with the 2.6 × 60 m snow fence, providing a proxy for climate change. Results from the first year of operation in 1994 showed that the period of continuous snow cover was increased by 90 d. The deeper and earlier snowpack behind the fence insulated soils from winter air temperatures, resulting in a 9°C increase in annual minimum temperature at the soil surface. The extended period of snow cover resulted in subnivial microbial activity playing a major role in annual C and N cycling. The amount of C mineralized under the snow as measured by CO₂ production was 22 g m^-2 in 1993 and 35 g m^-2 in 1994, accounting for 20% of annual net primary above-ground production before construction of the snow fence in 1993 and 31% after to snow fence was constructed in 1994. In a similar fashion, maximum subnivial N₂O flux increased 3-fold behind the snow fence, from 75 μg N m^-2 d^-1 in 1993 to 250 μg N m^-2 d^-1 in 1994. The amount of N lost from denitrification was greater than the annual atmospheric input of N in snowfall. Surface litter decomposition studies show that there was a significant increase in the litter mass loss under deep and early snow, with no significant change under medium and little snow conditions. Changes in climate that result in differences in snow duration, depth, and extent may therefore produce large changes in the C and N soil dynamics of alpine ecosystems.