Water shortage is a serious problem threatening sustainable development of agriculture in the North China Plain, where winter wheat is the largest water-consuming crop. The objective of this study was to optimize irrigation scheme for high yield and high water use efficiency (WUE) in wheat (Triticum aestivum L.), on the basis of Jimai 22, a represented cultivar in production. In the field experiments conducted in growing seasons, unfixed amount of water was supplied at sowing, jointing, and anthesis stages to adjust the soil moisture into a controlled ladder. For example, the relative soil moisture contents in the W0 treatment were 85% at sowing, 80% at jointing, and 70% at anthesis; in the W1 treatment, they were 85%, 85%, and 70%, respectively; analogically, they were 85%, 85%, and 75% in the W2 treatment and 85%, 90%, and 80% in the W3 treatment. The results showed that WUE was higher in W0 than in other treatments. However, the grain yield was the lowest in treatment W0. In growing seasons, the grain yield, irrigation water use efficiency (WUEI), precipitation use efficiency (WUEP), and irrigation benefit (IB) in W1, W2, and W3 were decreased significantly because more water was supplied. Under the experimental condition, the W1 regime was considered as the optimum. In this regime, the relative soil moisture contents at 0–140 cm soil layer were controlled to 85% at sowing, 80% at jointing, and 70% at anthesis stages. When 15.32 mm of water was supplied in growing seasons, the final grain yields of W1 treatment were 8185.75 kg ha-1, and the WUEI and WUEP were the highest among the 4 treatments. The deep soil water content (0-140 cm) can be estimated by shallow soil moisture content (0-60 cm), which is more conducive to estimate the deep soil water content through shallow soil and reduce the workload of supplemental irrigation based on testing soil moisture.