A pot experiment with wheat plants was carried out to study how late application of nitrogen (N) fertilizer affects the use of pre-anthesis N reserves during the grain-filling period. Increasing doses of N fertilizer were applied (0, 40, and 52 mg N plant1), either in two amendments (growth stages GS20 and GS30, according to Zadoks scale) or in three amendments (GS20, GS30, and GS37). The experiment was arranged in a complete randomized three-block design with 129 plants per treatment. The plants were watered daily, harvested every 2 d between anthesis and maturity, and were separated into roots, leaf sheaths, leaf blades, and ears for further N determination. Grain N concentration improved due to a late N application in GS37 by 14% (higher N dose) and by 7% (further splitting the same N-fertilizer dose, respectively). The higher the N-fertilizer dose applied, the greater was the amount of pre-anthesis reserves in vegetative organs, these reserves became later available for remobilization. Although splitting the same N dose in three amendments did not increase the N reserves, these reserves were more efficiently remobilized allowing an improvement in grain N concentration. The fertilizer management did not change the temporary pattern of N accumulation in the ear, but did induce a change in the amount of N remobilized and in the contribution of each organ (root, leaf sheath, leaf blade) to this remobilization. Late N amendment allowed a greater N availability of leaf blades and ear N reserves (from 20% up to 26% and from 19% up to 22%, respectively) for remobilization towards the grain, decreasing the root contribution from 28% down to 15%, while the contribution of leaf sheaths was maintained around 35% irrespective of the N applied.