Empirical and modelled data from a model investigation into the consequences of nitrogen (N) deposition and nutrient manipulation on carbon (C) and nutrient cycling in phosphorus (P)-limited grasslands. Empirical data show above-ground biomass C, soil organic C and total soil N from two grassland types at Wardlow Hay Cop in the Peak District national park, UK. Wardlow is a long-term nutrient manipulation experiment (> 25 years) investigating the consequences of N deposition on grassland ecosystems. These data were collected during the summer of 2019 and were combined with total soil P data collected previously to form a dataset for inclusion in a CNP biogeochemical cycling model; N14CP. We use these empirical data to drive and calibrate the N14CP model in order to develop our understanding of the C, N and P dynamics of the two grasslands. Specifically, we investigate how potential differences in organic P cycling between the two grassland types may explain their contrasting responses to long-term N deposition and further experimental treatments. Accordingly, the bulk of this dataset is modelled data derived from the empirical data, and relates to the responses of plant C, soil C, N and P to N deposition and nutrient manipulation. This includes data on the CNP budgets of the modelled grasslands, P-cycling parameters used within the model, comparisons of empirical to modelled data, and changes in CNP pools resulting from N deposition and nutrient manipulation.