Abstract Background and aims Cover crops can efficiently increase nitrogen (N) recycling in agroecosystems. By providing a green-manure effect for the next crop, they allow reduced mineral fertilisation. We developed a decision-support tool, called MERCI, to predict N available from cover crop residues over time, from a single measurement of fresh shoot biomass. Methods We coupled a large experimental database from France with a simulation experiment using the soil-crop model STICS. More than 25 000 measurements of 74 species of cover crops as a sole crop or bispecific mixtures were collected. Linear regression models, at the species, family or entire-database level depending on the data available, were built to predict dry biomass, N amount and C:N ratio. Dynamics of N mineralized and leaching from cover crop residues were predicted at 24 contrasting sites as a function of the biomass, carbon (C):N ratio and termination date. Results Correlations between fresh biomass, dry biomass and N amounts in experimental data were strong (r = 0.80-0.96), and predicted N amounts in fresh shoot biomass were relatively accurate. Percentages of N mineralized and leached simulated by STICS were explained mainly by the C:N ratio, site and number of months after termination, but to different degrees. Conclusion MERCI is an easy and robust decision-support tool for predicting N release in the field, and could thus be adopted by advisors and farmers to improve management of nutrient recycling in temperate arable cropping systems.