Atmospheric nitrogen (N) deposition is a global and increasing threat to biodiversity and ecosystem function. Much of our current understanding of N deposition impacts comes from field manipulation studies, although interpretation may need caution where simulations of N deposition (in terms of dose, application rate and N form) have limited realism. Here, we review responses to simulated N deposition from the UKREATE network, a group of nine experimental sites across the UK in a diversity of heathland, grassland, bog and dune ecosystems which include studies with a high level of realism and where many are also the longest running globally on their ecosystem type. Clear responses were seen across the sites with the greatest sensitivity shown in cover and species richness of bryophytes and lichens. Productivity was also increased at sites where N was the limiting nutrient, while flowering also showed high sensitivity, with increases and declines seen in dominant shrub and forb species, respectively. Critically, these parameters were responsive to some of the lowest additional loadings of N (7.7–10 kg ha−1 yr−1) showing potential for impacts by deposition rates seen in even remote and ‘unpolluted’ regions of Europe. Other parameters were less sensitive, but nevertheless showed response to higher doses. These included increases in soil %N and ‘plant available’ KCl extractable N, N cycling rates and acid–base status. Furthermore, an analysis of accumulated dose that quantified response against the total N input over time suggested that N impacts can ‘build up’ within an ecosystem such that even relatively low N deposition rates can result in ecological responses if continued for long enough. Given the responses have important implications for ecosystem structure, function, and recovery from N loading, the clear evidence for impacts at relatively low N deposition rates across a wide range of habitats is of considerable concern.