The effect of snow cover on surfaceatmosphere exchanges of nitrogen oxides (nitrogen oxide (NO) ? nitrogen dioxide (NO 2 ); note, here 'NO 2 ' is used as surrogate for a series of oxidized nitrogen gases that were detected by the used monitor in this analysis mode) was investigated at the high elevation, subalpine (3,340 m asl) Soddie site, at Niwot Ridge, Colorado. Vertical (NO ? NO 2 ) concentration gradient measurements in interstitial air in the deep (up to *2.5 m) snowpack were conducted with an automated sampling and analysis system that allowed for continuous observations throughout the snow-covered season. These measurements revealed sustained, highly elevated (NO ? NO 2 ) mixing ratios inside the snow. Nitrogen oxide concentrations were highest at the bottom of the snowpack, reaching levels of up to 15 ppbv during mid-winter. Decreasing mixing ratios with increasing distance from the soilsnow interface were indicative of an upwards flux of NO from the soil through the snowpack, and out of the snow into the atmosphere, and imply that biogeochemical processes in the subnival soil are the predominant NO source. Nitrogen dioxide reached maximum levels of *3 ppbv in the upper layers of the snowpack, i.e., *20-40 cm below the surface. This behavior suggests that a significant fraction of NO is converted to NO 2 during its diffusive transport through the snowpack. Ozone showed the opposite behavior, with rapidly declining levels below the snow surface. The mirroring of vertical profiles of ozone and the NO 2 /(NO ? NO 2 ) ratio suggest that titration of ozone by NO in the snowpack contributes to the ozone reaction in the snow and to the ozone surface deposition flux. However, this surface efflux of (NO ? NO 2 ) can only account for a minor fraction of ozone deposition flux over snow that has been reported at other mid-latitude sites. Neither (NO ? NO 2 ) nor ozone levels in the interstitial air showed a clear dependence on incident solar irradiance, much in contrast to observations in polar snow. Comparisons with findings from polar snow studies reveal a much different (NO ? NO 2 ) and ozone snow chemistry in this alpine environment. Snowpack concentration gradients and diffusion theory were applied to estimate an average, wintertime (NO ? NO 2 ) flux of 0.005-0.008 nmol m -2 s -1 , which is of similar magnitude as reported (NO ? NO 2 ) fluxes