Abstract: Plastic mulching technology is widely used for agricultural production to limit crop evapotranspiration. In order to clarify the potential of plastic mulching on agricultural productivity and to explore the effect of maximum water-saving on the growth and yield of winter wheat in arid/semi-arid regions, a field experiment was conducted at the National Experiment Station of Precision Agriculture in Changping, Beijing (N40°10′33.26″, E116°23′37.07″). The experiment consisted of four treatments — T1 (plastic film mulching with no irrigation), T2 (wintering watering with no film mulching), T3 (wintering and jointing watering with no film mulching), and T4 (wintering, jointing and flowering watering with no film mulching). Compared with conventional film mulching, T1 treatment was covered with 1 cm soil layer over plastic film surface. Water consumption characteristics and yield were determined using the weighting lysimeter over the whole growth stage and sampling method. Also concurrently, the 200 cm depth soil moisture content was measured (using Diviner 2000) every day for the period from April to May. The results showed that T1 efficiently utilized soil moisture. There was a cubic function between cumulative evapotranspiration and the number of days after sowing for the four treatments, with determination correlation coefficient R² > 0.99. The heading stage had the maximum theoretical and actual values of soil-crop coefficient (Kc) under T1 and T4. For T2 and T3, Kc had a quadratic function with the number of days after sowing during jointing stage, with a determination correlation coefficient R² > 0.70 (that for T2 was 0.69). Water consumption was significantly lower under T1 than under others treatments during sowingjointing period. However, during jointingmaturity period, T1 and T2 were not statistically different in terms of water consumption, but had lower value than T3 and T4 (P < 0.05). During bootingflowering period and floweringmaturity period, water consumption under T1 was higher, respectively, by 3.1 mm and 21.43 mm than that of T2, but not statistically different (P > 0.05). However, this increased water consumption at the 50–100 cm soil depth at late growth stage. The peak evapotranspiration under T1 was higher than that of T2, but both lower than those of T3 and T4. The time of peak evapotranspiration under T1 (215 d after sowing) came later than those of T2, T3 and T4 (194 d after sowing). The time of Kc-max of T1 was consistent with that of T4 (214 d after sowing), but occurred later than that of T2 (200 d after sowing) and T3 (199 d after sowing). Compared with T2 and T3, T1 enhanced flag leaf water potential and delayed leaf-senescence, with soil surface (0–5 cm layer) temperature improving by 0.5 ℃. This temperature increase was not statistically significant (P > 0.05) and contributed minimally to soil evaporation. In terms of yield, yield components and water use efficiency, grain number per spike and the 1000-grain weight of T1 were higher than those of T2 and T3, which were in turn lower than those of T4; but were not statistically different (P > 0.05). The yield of T1 was not statistically different from that of T2 and T3, but significantly lower than that of T4 (P < 0.05) and with 22.6% increase in water use efficiency (P < 0.05). It suggested that plastic film mulching could widely replace wintering and jointing water application by limiting soil evaporation at the early growth stage in semi-humid regions with sufficient soil water content before sowing. The water saved was used for the late growth stage, which thereby reduced the amount of water consumption and increased water use efficiency of winter wheat. Irrespectively, this form of water-saving was not at the expense of crop yield.