Availability of nitrogen and water are major constraints for crop yield, and their interactions are manyfold. Yield gap analysis in rainfed systems commonly uses water-limited yield potential (Yw) as a benchmark; benchmarking against nitrogen-limited yield potential (YN) is less common. The aim of this study was to benchmark wheat yield against YN in winter-rainfall regions of Australia. We established experiments, and sampled both farmers’ fields and National Variety Trials spanning wide ranges of soils, management practices, and water supply (seasonal rainfall + irrigation) from 153 to 759 mm in South Australia, and from 178 to 428 mm in Western Australia. We measured yield, quantified crop nitrogen nutrition index (NNI) at stem elongation, anthesis or both, derived boundary functions relating YN and NNI, calculated yield gaps as the difference between YN and actual yield, and explored the associations between yield gaps and environmental, crop and agronomic factors.In South Australia, NNI at anthesis ranged from 0.45 to 1.45 and yield from 0.9 to 8.9 t ha−1; in Western Australia NNI at stem elongation ranged from 0.27 to 1.16 and yield from 1.0 to 7.1 t ha−1. Bi-linear boundary functions were fitted with a linear YN – NNI phase up to a NNIx threshold (0.95 ± 0.134 in South Australia, 0.87 ± 0.145 in Western Australia), and a plateau reflecting yield potential (7.8 ± 0.38 t ha−1 in South Australia, 6.5 ± 0.52 t ha−1 in Western Australia). Similar bi-linear boundaries, with congruent NNIx (≈0.9), were found for grain number and shoot biomass at maturity. In South Australia, water supply explained 54% of the yield gap, which declined linearly from about 6 t ha−1 to zero with increasing water availability. Further, the yield gap correlated negatively with carbon isotope discrimination at anthesis, a direct measure of crop water status. In Western Australia, the direct association between yield gap and rainfall was weak, but there was an indirect agronomic link, where low seeding rate in low rainfall environments contributed to low biomass at stem elongation, which in turn explained a large part of the yield gap.In two out of five National Variety Trials, where the aim is comparing the yield of current and emerging varieties, crops were nitrogen deficient. This is a potential source of bias as some varieties were above and others below the NNIx threshold of nitrogen sufficiency. Supply of nitrogen in varietal comparisons needs attention.The approach advanced in this paper can be applied to benchmark yield against crop nitrogen status and identify causes of yield gaps, and for in-season quantification of crop nitrogen status to assist fertiliser decisions directly, or indirectly as a reference for spectral indices.