Nitrogen fertilizer management in rainfed Mediterranean environments can be nancially risky because of the strong interaction between N and water availability on yield. is study was conducted to investigate whether the use of split-applications of N fertilizer that targeted speci c growth stages could improve grain yield, grain protein concentration (GPC), and nitrogen use e ciency (NUE) of dryland wheat (Triticum aestivum L.). Experiments with 7 N-application timings, two N-rates, and two wheat varieties were conducted at two sites over two seasons. Despite the seasonal rainfall in both years being below the historic averages, delayed or split N applications were able to signi cantly increase grain yield (2.50 vs. 2.25 t ha -1 ), GPC (14.0 vs. 12.4%), N uptake (111.2 vs. 101.2 kg ha -1 ), N harvest index (70.7 vs. 61.7%), agronomic e ciency (4.6 vs. 1.7 kg kg -1 ), and apparent recovery (59.3 vs. 44.6%) compared with N applied only at sowing. No single N treatment was able to consistently produce the highest yield, GPC or NUE across the N rates, locations, and years because of the in uence of timing and amount of rainfall during the season. Compared with applying all the N at sowing a targeted use of N, using split/delayed N applications at speci c growth stages to manage canopy development in response to the seasonal pattern of rainfall, increased yield, quality, and NUE, even in years with below average rainfall, thus lessening exposure to nancial risk with rainfed wheat in dry years.