Nitrogen (N) fixation, denitrification, and ecosystem pools of nitrogen were measured in three subarctic ecosystem types differing in soil frost-heaving activity and vegetation cover. N-2-fixation was measured by the acetylene reduction assay and convened to absolute N ecosystem input by estimates of conversion factors between acetylene reduction and N-15 incorporation. One aim was to relate nitrogen fluxes and nitrogen pools to the mosaic of ecosystem types of different stability common in areas of soil frost movements. A second aim was to identify abiotic controls on N2-fixation by simultaneous measurements of temperature, light, and soil moisture. Nitrogen fixation rate was high with seasonal input estimated at 1.1 g N m(-2) on frost-heaved sorted circles, which was higher than the total plant N content and exceeded estimated annual plant N uptake several-fold but was lower than the microbial N content. Seasonal fixation decreased to 0.88 g N m(-2) on frost-heaved moss-covered surfaces and to 0.25 g N m(-2) in stable heath vegetation, both lower than the plant and microbial N content. Yet fixation was estimated to correspond to about 2.7 times the annual plant N demand on the moss-covered surfaces but less than the plants’ demand on the heath. Surprisingly, we found no denitrification on any surface. Climatic changes in the Arctic will generate a warmer climate and change precipitation patterns. A warmer, drier environment will decrease N2-fixation and thereby N availability to plants and microorganisms, while wetter conditions probably will increase N2-fixation and thereby N supply to the surroundings.