Abstract: Based on the characteristics of the paddy soil nutrient loss were mainly concentrated in the early stage of rice growth in Taihu area. A field experiment was conducted to study the effects of postponing N application on rice yield and nitrogen runoff loss. The super rice cultivar of Wuyunjing 29 was field-grown. Four nitrogen application ratios such as basal, tillering fertilizer and panicle fertilizer 5: 2: 3 (O1), 4: 2: 4 (O2), 3: 2: 5 (O3), 2: 2: 6 (O4) were designed, in the amount of nitrogen applied to 300 kg/hm2, with deep basal fertilizer machinery. The result showed that: for the yield components of rice, compared to the conventional nitrogen application (O2), postponing N application (O4) significantly decreased the effective panicles and spikelets number per panicle, but filled grain percentage and 1 000-grain weight showed the opposite trends. Finally O4 treatment of postponing N application declined rice yield a little, but there was no significant difference. There was six times of farmland overland runoff in rice season, and the amount of runoff was 3 567.0 m3/hm2 in this research season. With the increase of the proportion of postponing N application, the total nitrogen runoff loss decreased. The total N loss of O4 treatment was significantly decreased by 12.02% compared with that of O2 treatment. At the same time, postponing N application could decrease the total nitrogen loss rate with overland runoff in the paddy field and nitrogen drift loss rate of rice.