Nitrogen (N) fixation in moss-associated cyanobacteria is one of the main sources of available N for N-limited ecosystems such as subarctic tundra. Yet, N2 fixation in mosses is strongly influenced by soil moisture and temperature. Thus, temporal scaling up of low-frequency in situ measurements to several weeks, months or even the entire growing season without taking into account changes in abiotic conditions cannot capture the variation in moss-associated N2 fixation. We therefore aimed to estimate moss-associated N2 fixation throughout the snow-free period in subarctic tundra in field experiments simulating climate change: willow (Salix myrsinifolia) and birch (Betula pubescens spp. tortuosa) litter addition, and warming. To achieve this, we established relationships between measured in situ N2 fixation rates and soil moisture and soil temperature and used high-resolution measurements of soil moisture and soil temperature (hourly from May to October) to model N2 fixation. The modelled N2 fixation rates were highest in the warmed (2.8 ± 0.3 kg N ha−1) and birch litter addition plots (2.8 ± 0.2 kg N ha−1), and lowest in the plots receiving willow litter (1.6 ± 0.2 kg N ha−1). The control plots had intermediate rates (2.2 ± 0.2 kg N ha−1). Further, N2 fixation was highest during the summer in the warmed plots, but was lowest in the litter addition plots during the same period. The temperature and moisture dependence of N2 fixation was different between the climate change treatments, indicating a shift in the N2 fixer community. Our findings, using a combined empirical and modelling approach, suggest that a longer snow-free period and increased temperatures in a future climate will likely lead to higher N2 fixation rates in mosses. Yet, the consequences of increased litter fall on moss-associated N2 fixation due to shrub expansion in the Arctic will depend on the shrub species’ litter traits.