The feasibility of using legume leys to redress declining levels of soil nitrogen (N) fertility on the heavy clay Vertisols of the northern Australian grain belt depends partly on the ability of plant residues to supply N directly to subsequent cereal crops. An alternative is the use of fertiliser N in continuous cereal cropping. Two experiments were conducted (one in the field, the other under polyhouse conditions) to compare the uptake of N from either plant residues or ammonium sulfate fertiliser that had been labelled with 15N. In a field trial, 15N-labelled shoots of grain sorghum and Desmanthus virgatus and ammonium sulfate were applied to micro-plots and the flux of the added N between different soil pools and a wheat crop was followed over 219 days. Only small amounts of residue-derived N (<5%) were recovered in the mineral N of the soil at a depth of 0-10 cm, whereas over 88% of the fertiliser N was present as mineral N soon after adding the fertiliser. Soil microbial biomass-N was increased following addition of residues. Recovery of added 15N in the wheat crop was much higher from the fertiliser (35%) than from the 2 residue sources (<5%). The pot trial compared a wider range of 15N-labelled residues (shoot and root residues of Desmanthus virgatus, Lablab purpureus, and sorghum) with several rates of ammonium sulfate, applied in the presence and absence of non-labelled grain sorghum residues, over 4 cropping cycles. Dry matter production and N uptake were increased by application of fertiliser N, although the response was reduced in the presence of non-labelled sorghum residues; responses to residue N were much smaller than those to fertiliser N. In the first crop following residue application <7% of residue N was recovered, increasing to 12-23% over the 4 crops. Recovery of fertiliser N by the crops increased with the rate of application, and also depended on whether it was applied together with residues. A feature of the results, in both the field and pot experiments, was the large proportion of applied 15N that could not be accounted for in either the soil or the crops, and these losses have been attributed to denitrification.