Nitrous oxide (N2O) is an important greenhouse gas. N2O emissions from soils vary with fertilization and cropping practices. The response of N2O emission to fertilization of agricultural soils plays an important role in global N2O emission. The objective of this study was to assess the seasonal pattern of N2O fluxes and the annual N2O emissions from a rain-fed winter wheat (Triticum aestivum L.) field in the Loess Plateau of China. A static flux chamber method was used to measure soil N2O fluxes from 2006 to 2008. The study included 5 treatments with 3 replications in a randomized complete block design. Prior to initiating N2O measurements the treatments had received the same fertilization for 22 years. The fertilizer treatments were unfertilized control (CK), manure (M), nitrogen (N), nitrogen + phosphorus (NP), and nitrogen + phosphorus + manure (NPM). Soil N2O fluxes in the highland winter wheat field were highly variable temporally and thus were fertilization dependent. The highest fluxes occurred in the warmer and wetter seasons. Relative to CK, m slightly increased N2O flux while N, NP and NPM treatments significantly increased N2O fluxes. The fertilizer induced increase in N2O flux occurred mainly in the first 30 days after fertilization. The increases were smaller in the relatively warm and dry year than in the cold and wet year. Combining phosphorous and/or manure with mineral N fertilizer partly offset the nitrogen fertilizer induced increase in N2O flux. N2O fluxes at the seedling stage were mainly controlled by nitrogen fertilization, while fluxes at other plant growth stages were influenced by plant and environmental conditions. The cumulative N2O emissions were always higher in the fertilized treatments than in the non-fertilized treatment (CK). Mineral and manure nitrogen fertilizer enhanced N2O emissions in wetter years compared to dryer years. Phosphorous fertilizer offset 0.50 and 1.26 kg N2O-N ha−1 increases, while manure + phosphorous offset 0.43 and 1.04 kg N2O-N ha−1 increases by N fertilizer for the two observation years. Our results suggested that the contribution of single N fertilizer on N2O emission was larger than that of NP and NPM and that manure and phosphorous had important roles in offsetting mineral N fertilizer induced N2O emissions. Relative to agricultural production and N2O emission, manure fertilization (M) should be recommended while single N fertilization (N) should be avoided for the highland winter wheat due to the higher biomass and grain yield and lower N2O flux and annual emission in m than in N.