Arid ecosystems cover similar to 40% of the Earth's terrestrial surface and store a high proportion of the global nitrogen (N) pool. They are low-productivity, low-biomass, and polyextreme ecosystems, i.e., with (hyper)arid and (hyper)oligotrophic conditions and high surface UV irradiation and evapotranspiration. These polyextreme conditions severely limit the presence of macrofauna and -flora and, particularly, the growth and productivity of plant species. Therefore, it is generally recognized that much of the primary production (including N-input processes) and nutrient biogeochemical cycling (particularly N cycling) in these ecosystems are microbially mediated. Consequently, we present a comprehensive survey of the current state of knowledge of biotic and abiotic N-cycling processes of edaphic (i.e., open soil, biological soil crust, or plant-associated rhizosphere and rhizosheath) and hypo/endolithic refuge niches from drylands in general, including hot, cold, and polar desert ecosystems. We particularly focused on the microbially mediated biological nitrogen fixation, N mineralization, assimilatory and dissimilatory nitrate reduction, and nitrification N-input processes and the denitrification and anaerobic ammonium oxidation (anammox) N-loss processes. We note that the application of modern meta-omics and related methods has generated comprehensive data sets on the abundance, diversity, and ecology of the different N-cycling microbial guilds. However, it is worth mentioning that microbial N-cycling data from important deserts (e.g., Sahara) and quantitative rate data on N transformation processes from various desert niches are lacking or sparse. Filling this knowledge gap is particularly important, as climate change models often lack data on microbial activity and environmental microbial N-cycling communities can be key actors of climate change by producing or consuming nitrous oxide (N2O), a potent greenhouse gas.