AbstractBackgroundA high use‐efficiency of fertilizer N remains essential to sustain high crop productivity with low environmental impact. However, little is known on the long‐term lability of mineral fertilizer N.AimsTo quantify crop uptake and leaching of 15N‐labelled mineral fertilizer that has been retained in an agricultural soil for 25–30 years in crops with variable growing season.MethodsA field plot received 15N‐labelled mineral fertilizers over a period of 5 years and was then kept under arable cropping for 12 years. After relocation to 16 lysimeters, the topsoil grew set‐aside grassland for the next 13 years. Then crop uptakes and leaching losses of 15N remaining in soil was tested over a 2‐year period by either converting set‐aside grass to production grassland, or by replacing it with spring barley (+/− autumn cover crop) or vegetation‐free fallow. All treatments received unlabelled mineral N fertilizers.ResultsCrop uptake and leaching of 15N were generally highest in the first test year after termination of the set‐aside. The leaching of residual 15N in soil declined in the order: vegetation‐free soil (4.7%), spring barley (1.9%), spring barley + cover crop (0.7%) and production grassland (0.2%). Corresponding losses for the second leaching period were 2.7%, 0.9%, 0.4% and 0.06%. There was a fixed relationship between leaching losses of 15N and total N.ConclusionsAfter residing in soil for 25–30 years, the lability of labelled mineral N fertilizer residues appeared slightly higher than the lability of bulk soil N. Autumn vegetation was crucial for reducing leaching losses.