Dual-purpose crops are grazed during the vegetative phase by cattle, goats and sheep in many dryland farming systems. After grazing, these crops continue to grow until maturity when they are harvested for grain. The process of grazing depletes crop nitrogen (N) content. Yield and grain protein concentration may therefore be more N-limited in grazed than in grain-only crops. This paper reviews 14 field experiments in south-eastern Australia, 12 of which were on farms, to measure the responses to N fertilizer by wheat, barley or canola crops defoliated by mowing or grazing with sheep. In five experimental comparisons of defoliated and grain-only crops, the average nitrogen use efficiency (NUE) of fertilizer N applied soon after defoliation, as is typically practiced, was 18 %, while NUE of grain-only crops fertilized at the same time and N-rate was 41 %. Applying N fertilizer to defoliated crops at this stage was usually unprofitable. In one experiment comparing responses of grazed and grain-only crops to time of N application post-grazing, a delay of two weeks after grazing increased yield of grain-only crops and NUE of both grazed and grain-only crops, compared with application soon after grazing. Our explanation of these results is the low N demand by small, defoliated crops and increased N demand as they regrew after defoliation. Additional evidence for the N-demand explanation came from a positive relationship between NUE of grazed crops and the yield of grain-only crops. The results suggest that delaying N application for two weeks after grazing and retaining more biomass after grazing can increase N demand, yield and NUE of grazed crops.