Combustion efficiency, as defined by the proportion of the burned biomass in the whole biomass, is the key parameter to assess the amount of greenhouse gas release from biomass burning. However, combustion efficiency depends on many different factors, and various results were reported by different researchers. In this paper, GIS and geostatistical methods are simultaneously used to calculate the spatial pattern and combustion efficiency of high intensity meadow fires, for typical meadow fire burned area selected in Daxinganling Mountains. CBERS-02B satellite image was examined to determine the study area and sample plots, by using the vegetation index and the threshold value extraction method. Five sample plots (20m?19m, sample plot 3 is 20m?20m) were thus determined where the combustion depth was measured every 1m distance and the coordinate of each measure point was recorded. At the same time, we selected 1m?1m unburned plots near the sample plots as reference, where the fuel depth and mass were measured. The study area was subject to high intensity fires, as evidenced by the fact that the surface fuel and litter were all burnt out. The mass of surface fuel and litter was calculated based on the reference sample plots. The humus layer was burnt into patches with different depths, which were measured based on the nearest points where humus was not burnt. The results show that the average fuel loading is 37.3t/ha, and grass layer, litter layer and humus layer account for 18.50%, 28.95% and 52.55% of the total fuel loading, respectively. The nugget coefficient falls within 80.84%-97.88%, and the burning depth has weak spatial correlation. The combustion efficiency is within 44.35%-90.6% for different fire intensities, leading to average value of 64.51%. The complexity of combustion processes hinders the calculation of combustion efficiency for large-scale forest fires. Due to the inner relation between fire danger and fire behavior, fire behavior influences combustion efficiency directly, and so it is a possible way to calculate the combustion efficiency for large-scale forest fires by analyzing the relation of combustion efficiency and fire danger, with other related factors such as fuel type and weather conditions included.