Ice core records of NO3− deposition to polar glaciers could provide unrivaled information on past photochemical status and N cycling dynamics of the troposphere, if the ice core records could be inverted to yield concentrations of reactive N oxides in the atmosphere at past times. Limited previous investigations at Summit, Greenland, have suggested that this inversion may be difficult, since the levels of HNO3 and aerosol-associated NO3− over the snow are very low in comparison with those of NO3− in the snow. In addition, it appears that some fraction of the NO3− in snow may be reemitted to the atmosphere after deposition. Here we report on extensive measurements of HNO3, including vertical gradients between 1.5 and 7 m above the snow, made during the summers of 1994 and 1995 at Summit. These HNO3 data are compared with NO3− concentrations in surface snow and the first measurements of the concentrations and fluxes of total reactive nitrogen oxides (Ny) on a polar glacier. Our results confirm that HNO3 concentrations are quite low (mean 0.5 nmol m−3) during the summer, while NO3− is the dominant ion in snow. Daytime peaks in HNO3− appear to be due at least partly to emissions from the snow, an assertion supported by gradients indicating a surface source for HNO3− on many days. Observed short-term increases in NO3− inventory in the snow can be too large to be readily attributed to deposition of HNO3− suggesting that deposition of one or more other N oxides must be considered. We found that the apparent fluxes of HNO3 and NOy were in opposite directions during about half the intervals when both were measured, with more cases of HNO3 leaving the snow, against an NOy flux into the snow, than the reverse. The concentrations of NOy are generally about 2 orders of magnitude greater than those of HNO3; hence deposition of only a small, non-HNO3, fraction of this pool could dominate NO3− in snow, if the depositing species converted to NO3−, either in the snowpack or upon melting for analysis.