Abstract: Agriculture and pasture ecotone are typical ecologically vulnerable areas. Crop productivities and the agricultural sustainability are greatly affected by climate change in these areas. Field experiments were conducted from 2011 to 2013 in Inner Mongolia (41°06′N, 111°28′E) to investigate the influence of planting dates on yield and phenology of spring wheat and potential of improving resource (water and nitrogen) use efficiencies. The experimental field was located in arid and semiarid climate zone, where mean annual temperature was 2.7℃, annual growing degree days (>0 ) was about 2553 (℃·d), the average annual rainfall was only 354 mm mainly in July and August accounting for 80% precipitation in the whole growing season. Five sowing dates was designed from 26 April to 5 June with a 10-day interval. The results showed that the yields after 6 May were significantly (p＜0.05) decreased due to delayed sowing dates. Compared to early sowing date (26 April), the yield of the late sowing date was reduced by 63.3% to 72.3% due to the obvious reduction in daily temperature from the growth stages of heading to maturity. Harvest index (HI) and grain weight per 1000 seeds were also affected by sowing dates. The early sowing (26 April and 5 May) treatments had higher HI. But after 16 May, the HI decreased significantly (p＜0.05). The grain weight per 1000 seeds decreased from 44.65 to 20.73 g with the delaying of the seeding dates, which had a trend similar with HI. It indicated that low values of HI and grain weight contributed to yield decrease when sowing after 26 May. Phenology of wheat as expressed in calendar days were different between years and sowing dates, especially from sowing to emergency, but the phenology based on physiological development time (PDT) was similar with a constant value of 95.3 days (the days required to complete whole development stage at optimal temperature condition) between years and sowing dates. Water use efficiency (WUE) was significantly (p＜0.05) decreased by delaying sowing dates. Compared with early sowing treatments, the WUE of late sowing dates were reduced by 68.8% to 74.3% probably due to the reduction of ratio of spike over above-ground dry matter in late sowing treatments (i.e. more vegetative growth and bigger canopy than early sowing treatments). Total nitrogen uptake by above-ground biomass ranged from 14.0 to 17.0 g/m2 for different sowing dates, however, the difference was not statistically significant (p>0.05). As the early sowing treatments showed higher (p＜0.05) yield and HI, nitrogen use efficiency (NUE), the transformation of nitrogen uptake to the economic yield was significantly (p＜0.05) higher in early sowing treatments than in late sowing treatments. Therefore, the best sowing date should be earlier than 6 May, and early sowing before 6 May could improve yield, water and nitrogen use efficiencies of wheat in the agriculture and pasture ecotone.