Most of the nitrogen (N) used by Australian crops is mineralised from the residues of previous crops and pastures. Net N mineralisation was studied in 2 field experiments in southern NSW, one comparing different residue-management and tillage systems during continuous cropping and the other comparing residues of annual and perennial pastures in a pasture–crop system. After 14 years of continuous cropping, soil total N concentration had decreased by 50%. Neither stubble retention nor direct drilling affected potential N mineralisation or the decrease in total N. However, soil mineral N in the field was greater after direct drilling than cultivation and greater after stubble retention than stubble burning. There were 2 reasons for the discrepancy. One was because retained stubble conserved soil water, leading to periods of increased mineralisation. The other was that direct drilling and stubble retention reduced growth and N uptake by crops. In contrast to the similar rates of potential mineralisation under different tillage and stubble systems, there were significant differences following different pasture species. In a 5-year study of a pasture–crop system we measured net mineralisation following annual pasture based on subterranean clover and perennial pasture based on lucerne and/or the grasses phalaris and cocksfoot. Mineralisation generally decreased with number of years after pasture removal. Previous lucerne pastures led to slow net mineralisation in the first year after removal, apparently because of immobilisation by high C : N residues. Mineralisation in soil containing perennial grass residues was the highest measured. This high rate may be due to redistribution of N to the topsoil by roots of perennial grasses. The comparison of continuous crop and pasture–crop systems showed that the decline in soil N supply was not prevented by direct drilling and stubble conservation, but N mineralisation was increased by pastures, particularly those containing perennial grasses.