The effects of fertilizers and water content on N2O emission were studied using the three most typical plantation soils. Soil incubations were performed and fertilization and water content treatments were designed. At 25 % of saturated water content(SWC), N2O emissions from the soil treated with urea, KNO3, (NH4)2 SO4 and KH2 PO4 were compared at application rates of 0, 100, 200, 300 and 500kg/hm2 . At 80% of SWC, similar experiments were carried out but at only one application rate(500 kg/hm2 ). N2O emissions at various water contents(20%, 35%, 50%, 65%, 80% and 100% of SWC) were studied. At low water content(25% of SWC), neither nitrogen However, at high water content(80 % SWC), the fertilizers resulted in much greater N2O emission irregardless of soil tested. The highest low level but steadily increased to a very high level for the soil treated with KNO3. High NO3- content was a basis of high level of N2O emission. N2O emission rates from soils peaked shortly after flooding, rapidly dropping to a very low level in soil from non-legume plantations, but lasting for a relatively long period in soil from legume plantations. When soil water content increased equaling to or higher than 65%, the accumulated N2O emission over a period of 13 d ranged from 20.21-29.78 mg/m2 for S. superba, 30.57-70.12 mg/m2 for P. elliottii and 300.89-430.51 mg/m2 for A. mangium. The critical water content was 50% of SWC, above which a high level of N2O emission could be expected, and below which very little N2O emissions were detected. The results suggest that, at low water content( ＜50% of SWC), the fertilization practice is safe with regard to N2O emissions, but at high water content( > 50% of SWC), nitrogen fertilizer in the form of nitrate could yield a 100-fold increase in N2O emissions. Legume plantations like A. mangium should be avoided in low lands which could easily suffer from flooding or poor drainage.