The response of UK semi-natural soil and plant communities to long-term inputs of pollutant nitrogen (N) is presently poorly understood. This study aims to investigate the effects of up to 8 years simulated pollutant N additions on plant nutrition and key below ground nutrient cycling processes in an upland heath, an acid grassland and a calcareous grassland. An additional series of plots was established in the grasslands that received short-term N and phosphorus (P) inputs which enabled investigation of NIP interactions and the determination of plant biomass. Soil inorganic N concentrations indicated that all of the sites had become 'N saturated', and that P limitation may be of increasing importance, particularly in the heathland and calcareous grassland. This was confirmed in the calcareous grassland where significant increases in above ground biomass were seen in only P treated plots, while in the acid grassland, there were no growth responses to nutrient additions. A field-based bioassay was developed which demonstrated that both short and long-term inputs of N significantly increased root surface phosphomonoesterase (PME) activity of Plantago lanceolata seedlings. A parallel microcosm-based bioassay revealed a significant relationship between root surface PME activity of Agrostis capillaris and extractable NfLt concentrations (r = 0.86) in the long-term acid grassland plots, indicating that this site may have reached or be approaching N saturation. In response to long-term N additions, soil microbial biomass carbon increased in the heathland, decreased in the acid grassland and remained constant in the calcareous grassland. Chloroform fumigation indicated that microbial biomass P decreased in the acid and calcareous grasslands, while microbial biomass N increased in the heathland and acid grassland in the N treated plots. Soil PME activity was highly sensitive to the N treatments. It increased at all sites and was significantly correlated with extractable inorganic N concentrations (r = 0.71) in the calcareous grassland, indicating close coupling between N saturation and P limitation. This relationship was also seen in the heathland where there was increased utilisation of monoester P sources in standard and customised BIOLOG plates and increased respiration rates in soils amended with organic P compounds. The implications of these results are discussed, with reference to the critical loads concept and to recent research by other workers in similar and contrasting environments.