We studied two different grassland fertiliser management regimes on sand and peat soils: above-ground application of a combination of organic N-rich slurry manure and solid cattle manure (SCM) vs. slit-injected, mineral N-rich slurry manure, whether or not supplemented with chemical fertiliser (non-SCM). Measurements of field N mineralisation as estimated from herbage N uptake in unfertilised plots were compared with (i) potential N mineralisation as determined from a standard laboratory soil incubation, (ii) the contribution of groups of soil organisms to N mineralisation based on production-ecological model calculations, and (iii) N mineralisation calculated according to the Dutch fertilisation recommendation for grasslands. Density and biomass of soil biota (bacteria, fungi, enchytraeids, microarthropods and earthworms) as well as net plant N-uptake were higher in the SCM input grasslands compared to the non-SCM input grasslands. The currently used method in Dutch fertilisation recommendations underestimated actual soil N supply capacity by, on average, 102 kg N ha-1 (202 vs. 304 kg ha-1 = 34%). The summed production-ecological model estimate for N mineralisation by bacteria, fungi, protozoa, and enchytraeids was 87–120% of the measured potential soil N mineralisation. Adding the modelled N mineralisation by earthworms to potential soil N mineralisation explained 98–107% of the measured herbage N uptake from soil. For all grasslands and soil biota groups together, the model estimated 105% of the measured net herbage N uptake from soil. Soil biota production-ecological modelling is a powerful tool to understand and predict N uptake in grassland, reflecting the effects of previous manure management and soil type. The results show that combining production ecological modelling to predict N supply with existing soil N tests using aerobic incubation methods, can add to a scientifically based improvement of the N fertilisation recommendations for production grasslands.