Abstract: A continuous automatic measurement system was set up to measure N2O emissions from a summer maize field in Beijing suburb in 2002. This system consisted of 9 static chamber units, a pneumatic unit, a sampling unit, a sample injection unit, a sample splitting unit and a gas chromatograph (GC) . The chambers were adjustable in height from 0.25 m for young plants to 2.25 m for fully developed maize plants with a covered soil area being 1.4 by 1.4 m (1.96 m2). N was top dressed in the three leaf stage (Jul 16, 2002) with a dose of 100 kg/hm2 N(as ammonium biocarbonate ) and in the ten-extended-leaf stage (Aug 14, 2002) with a dose of 200 kg/hm2 N (as urea) , separately. The diurnal, seasonal variation of N2O flux as well as the influences of sampling time and sampling frequencies on the simulated results of N2O emission were studied in the paper.A large seasonal variation of N2O emission was observed during the summer maize season. There were two emission peaks which took placed 2 weeks after N fertilizer applied in the three leaf stage and the ten-extended leaf stage accounting for 22.8% and 64.8% of the total emission in whole growth season, respectively. The diurnal variation of N2O showed that the maximum N2O flux rate occurred at 11:00- 14:00, and the minimum flux of N2O occurred at 2:00 - 5: 00, which was coinciding to the diurnal variation of soil temperature. The accumulated N2O emissions based on the different sampling time and sampling frequencies showed that the sampling time strongly affected the calculated N2O emission because of the diurnal variation of N2O emission; while sampling frequencies had on influence on simulated results in this study.