Soil respiration varies not only in time but also in space. These temporal and spatial variations increase uncertainties in soil respiration measurements. Understanding the spatial and temporal variation in soil respiration of a region is fundamental for an accurate estimation of CO2 emission in the region. Based on the field measurements of soil respiration and some other environmental factors such as soil temperature and soil water content at 39 sites in Taiyuan basin in July, September and November 2008, we analyzed heterogeneous soil respiration and its relation to other environmental factors using both traditional and geostatistics methods. The traditional statistical analysis (descriptive statistics) showed that the averages of soil respiration rate were (7.8±3.3), (8.9±4.0) and (2.1±1.3)μmol CO2 m-2s-1, respectively, for July, September and November. The coefficients of variation were 42%, 44% and 59%, respectively, indicating that spatial variation of soil respiration in this area was moderate. When semivariance analysis of geostatistics was used, the results showed the semivariograms of soil respiration rate could well be described by a spherical model. The C0/(C0+C) values for July, September and November were 0.004, 0.038 and 0.005 respectively for soil respiration, and 0011, 0003 and 0.003 respectively for soil moisture, indicating that spatial heterogeneities of both soil respiration and soil water content were resulted mainly by structural factors. The ranges of semivariogram function of the 3 measurements were 190, 370 and 510 m respectively for soil respiration, and 90, 140 and 220 m respectively for soil moisture. The results indicated that the ecological progresses were influencing soil respiration and soil moisture which changed with time at different scales. The fractal dimension values of soil respiration for the measurements of July, September and November were 0.634, 0.965 and 0.763, respectively, indicating that the strongest spatial dependence of soil respiration is in July followed by November and the weakest is in September, in good agreement with the semi-variogram analysis.. The isoline map of soil respiration rates produced with the Kriging interpolation method showed that there was high spatial heterogeneity of soil respiration which was mainly caused by the differences in soil water content.