The experiment was conducted to investigate the effects of different forms of nitrogen application (ammonium, NH4(+) -N; nitrate, NO3(-) -N; and amide-N, NH2-N) and different concentrations (40, 200 and 800 mg/L) on N2O flux from the fluvo-aquic soil in a lab-simulated freezing and thawing process. N2O flux slowly decreased with soil freezing, and then increased slowly with soil thawing, reaching a peak at the initial thawing stage. The average N2O fluxes with addition of NH4(+) -N, NO3(-) -N and NH2 -N are 119.01, 611.61 and 148.22 microg x (m2 x h) (-1) respectively at the concentration of 40 mg/L; 205.28, 1084.40 and 106.13 microg x (m2 x h)(-1) at the concentration of 200 mg/L; 693.95, 1 820.02 and 49.74 microg x (m2 x h)(-1) at the concentration of 800 mg/L. The control is only 100.35 microg x (m2 x h)(-1). N2O flux with addition of NH4(+) -N and NO3(-) -N increased with increasing concentration, ranging from 17.49% to 425.67% for NH4(+) -N, and 563.38% to 1 458.6% for NO3(-) -N compared with control. There was a time lag achieving stable N2O emission flux with concentration increase. In contrast, by adding NH2 -N to soil, N2O flux decreased with increasing concentration. In sum, inorganic nitrogen such as NH4(+) -N or NO3(-) -N fertilizer incorporated in soil enhanced the cumulative N2O flux from the fluvo-aquic soil relative to amide-N. This study suggested that ammonium and nitrate concentration in overwintering water should be less than 200 mg/L and 40 mg/L in order to reduce N2O emissions from soil, regardless of amide-N.