A two-years experiment was designed to examine effects of different nitrogen (N) reduction and controlled-release urea application ratio on spring maize yield, water use efficiency and nitrate-N residue, so as to provide a reasonable fertilization management model for the dryland farming systems on the Loess Plateau. The experiment was performed from April 2017 to September 2018 in the rain-fed agricultural area of the Loess Plateau. Spring maize was planted with half plastic film mulching in one crop per annum. The experiment included six treatments, CK (no nitrogen fertilization treatment), N1C1 (the treatment of 65% controlled-release urea and 35% urea, N 200 kg/hm²), N1C2 (the treatment of 50% controlled-release urea and 50% urea, N 200 kg/hm²), N1C3 (the treatment of 35% controlled-release urea and 65% urea, N 200 kg/hm²), N1 (reduce nitrogen fertilization mode, urea, N 200 kg/hm²), N2 (conventional nitrogen fertilization mode, urea, N 250 kg/hm²).The soil moisture content, the yield of spring maize and nitrate nitrogen content in soil profile (0-300 cm) at harvest period were measured. The results of two years experiment showed that compared with the N2 treatment, N1 treatment did not reduced the crop yield, but significantly increased the crop yield (p<0.05), which increased by 9.6% and 6.9% respectively and significantly increased the water use efficiency (p<0.05), by 13.3% and 10.2% respectively in 2017 and 2018. Under the same nitrogen application rate (200 kg/hm²), compared with N1, the treatment of applying a certain proportion of controlled-release fertilizer reduced the yield of spring maize and the water use efficiency in 2017; the treatment of N1C2 significantly increased the yield of spring maize and the water use efficiency (p<0.05) by 7.7% and 11.6%, respectively in 2018. In addition, after two years of experiment, the reducing nitrogen application and applying a certain proportion of controlled-release fertilizer could significantly reduce the amount of residual nitrate nitrogen in soil profile (0-300 cm) (p<0.05), the N1 treatment reduced by 61.2% compared with the N2 treatment. Under the same nitrogen application rate (200 kg/hm²), the N1C2 treatment reduced by 50.8% compared with the N1 treatment.