The objective of this study was to evaluate N2O emissions during the first three months after different nitrogen (N) source treatments involving green manure and chemical fertilizer in a sugarcane field in Okinawa, Japan. Four treatment plots were established: 100 and 300 kg N ha−1 chemical fertilizer (100N and 300N plots, respectively); green manure (Crotalaria juncea, 150 kg N ha−1, C:N=20; GM plot); and the control (0N plot). We regularly measured N2O flux and soil inorganic N (NH4+-N and NO₃--N), and also recorded environmental data. There were no significant differences in the fluctuation pattern and peaks of N2O emissions between the 100N and GM plots, and total N2O emissions were comparable, though the applied N amount was higher in the GM plot. Little rainfall in the early period should limit microbial activity in the 100N plot, though soil NH4+-N and NO₃--N contents were higher in the 100N plot than in the GM plot. Total N2O emissions in the 300N plot were 1.5-times higher than in the 100N and GM plots. Collectively, the results suggest that the amount of applied N more strongly affected N2O emission than the difference in N substrate quality.