Soil nitrogen mineralization of the dominant vegetation types in Changbai Mountain, Northeast China, was simulated based on continuously observed soil temperature and published experiment data. Significant errors and biases, which were caused by unreasonable calculation and parameterization, in previous reports were verified. In this study, the nitrogen mineralization rates in different vegetation zones were estimated with temperature\|dependent regression models. The mineralization rates were 58.5 kg?hm-2?a-1 in the conifer\|broad leaf mixed forest which distributed in the foot of the mountain, 34.5 kg?hm-2?a-1 in the pine\|spruce\|fir mixed forest of subalpine, 28.5 kg?hm-2?a-1 in birch\|spruce\|fir mixed forest, and 19.6 kg?hm-2?a-1 in tundra. Based on the continuous records of soil temperature (t), an integrated regression model for estimating nitrogen mineralization rate (y, kg?hm-2?d-1), y=0.038255e(0.1392t), was established. The seasonal and vertical variation of nitrogen mineralization were also estimated with the regression model. The nitrogen mineralization was estimated to decline at a rate of 2.31 kg?hm-2?a-1 per 100m with elevation increase.