The observed increase in levels of atmospheric CO₂ and the signing of the UN Framework Convention on Climate Change have increased interest in the global warming and the global carbon cycle. As one of the key components in the global carbon cycle, the response of soil carbon pool to the global warming is another studied focus issue. Changes in soil CO₂ emission along the elevation gradient could reflect the outcome of multiple interacting environmental factors on a long temporal scale and a small spatial scale, therefore, a research focused on the variation of soil CO₂ emission along 1800-2155m elevation gradient and the response of it to controlling factors was carried out. In the study area, soil CO₂ emission increased from (1.94±0.06) μ mol m^-2 s^-1 to (2.22±0.07) μ mol m^-2 s^-1 along the elevation gradient, which was due to the responses of the controlling factors to the elevation. However, not only the effect of controlling factors can interact, overlap, or restrain each other, but also each factor can act in different ways. Soil temperature, soil moisture, SOC, total N and total P significantly correlated to soil CO₂ emission positively (n=14, P<0.05), and soil bulk density correlated to soil CO₂ emission negatively (n=14, P<0.05), whereas pH did not control soil CO₂ emission significantly in this study. Moreover, a variety of statistical approaches were used to characterize the primary controls over soil CO₂ emission among these controlling factors. Primary component analysis (PCA) explored data reduction and a stepwise regression analysis constructed a most parsimonious model of soil CO₂ emission, in which the most obvious effect of soil moisture appeared to be as a primary constraint on the distribution of soil CO₂ emission along the elevation gradient.