Current management of the worlds' grazing lands in either based on changes in plant species composition or on other management evaluation programs that emphasize changes in net aboveground production. Management is based solely on changes in aboveground production has been criticized as too limited in view, because it ignores root production, nitrogen pools, nutrient processes, and the long-term sustainability of the ecosystem. The purpose of this study was to compare the effects of elk (Cervus elaphus) grazing on aboveground production, internal nitrogen (N) fluxes, N pools and inputs, and elk nutrient transfers across the landscape in different vegetation types in Rocky Mountain National Park (RMNP), Colorado. Nitrogen processes and possibly N pools were significantly reduced in the willow community, but not in the upland grass/shrub community. Nitrogen mineralization rates were lower in grazed versus ungrazed short willow sites (P = 0.07; n = 4 sites), as were nitrate (NO3) pools (P = 0.10), but not in tall willow sites (P > 1.10 n = 4 sites) after 4 years. There was about half the annual N inputs to the soil surface in grazed willow sites (5.79g N/m²/yr = annual herbaceous biomass a?? offtake + litterfall + elk urine and feces) compared to ungrazed sites (9.66 g N/m²/yr = annual herbaceous biomass + litterfall), suggesting elk herbivory and movement led to a net loss of N in the willow vegetation type. Elk substantially reduced the annual growth of willows (Salix spp.) by 98% after 35 years and 66% after 4 years of treatment. Thus, height and canopy and N yield of willows were reduced as well as willow litter biomass (65 g/m2/yr in ungrazed versus 33 g/m2/yr in grazed), and N yield of willows was 64% less in grazed plots. Elk grazing had no significant effect on other soil N pools (NH4) or litter decomposition rates in either of the two willow types, nor on any nitrogen process rates or pools in the upland grass/shrub type (P > 0.10). Nitrogen concentrations in plant tissue were not influenced for the most part by elk grazing (increased N concentration found in only 4 of 13 species). Elk apparently also transported N away from aspen at an even higher rate (N inputs were 1.65 g N/m2/yr in grazed sites vs. 3.79 g N/m²/yr ungrazed). CENTURY modeling supported these observations. This soil N model predicted almost no change (0.5$ - 2% less) in N or carbon (C) pools in 50 years in the upland grass/shrub vegetation type, but greater losses in the willow type. If elk population levels were increased to carrying capacity in the ecosystem (about 25% more consumption) projected losses after 50 years were greater, but still <4% for soil C and 1% for spoil N in the upland type, and 6% for soil C and 2% for soil N in the willow type. Total shrub C was projected to decline 10% in 50 years in the willow type at current elk densities. We recommend conservative management of elk numbers and grazing until additional years of measures are gathered on the lower mineralization rates in the short willow type and until the projected declines in N pools can be verified.