Soil temperature regime refers generally to soil temperature at 50 cm (T50) in soil depth or at the interface between lithic or paralithicsubstance and soil in soil profiles thinner than 50 cm. It is an importantparameter charactering soil properties and a diagnostic feature and basis in modern soil classification systems for dividing soil classification units. However, so far none of the existing observatory stations iscapable of providingobservatory data all by itself sufficient for dividing attributes of soil temperature. Fortunately, as soil temperature at any observatory stationis closely related to geographic position (coordinates, altitude, etc.), meteorological conditions and some other environmental elements of the locality the station sits in, some researchers have built up several models by makinguse of these obtainable environmental indexes for assessing soil temperature regime. Currently the following five methods are commonly used in researches on soil temperature regime in China: “Soil temperature interpolation method”, “Direct estimation method”, “Newhall model-based estimation”, “National air temperature-based regression estimation”and “Coordinates and elevation-based regression estimation”. These methods have been extensively applied to researches on soil temperature regime in various regions of the country with some good results. However, little has been reported on the work In GuizhouProvince. In this paper, based on the ground climatic data (1951/71~1980) pooled from 86 meteorological observatory stations all over GuizhouProvince, the five methods were either used directly or modified in the light of the actual situation in Guizhou. Withthe soil temperature interpolation method, T50 is deduced from the soil temperatures measured at 40 cm and 80 cm in depth, with interpolation, namely, mean annual T50 =mean annual soil temperature at 40cm+[mean annual soil temperature at 80 cm-mean annual soil temperature at 40 cm]/4. Withthe direct estimation method, the mean annual soil temperature at 20 cm in depth is deemed as T50. Withthe Newhall simulation model method, some modification has been made of the model in the light of the close relationship between soil temperature and air temperature in Guizhou, likeassumingthat the mean annual T50 is 1.8 ℃ higher than the mean annual air temperature. The air temperature-based regression estimation method is developed by Feng Xueming, using the mean annual air temperaturesof 150 meteorological stations scattered in 30 provinces of China as (x) and mean annual T50(y), namely "y=2.9001+0.9513x" (r=0.9889).On the basis of the Feng Xuemingmethod, a multiple regression equation is put forward and used to reflect the influence of latitude (x_1) and elevation(x_2)besides the air temperature, namely "y=40.25-0.7166" x_1 "-" 0.002389x_2(r=0.9515). Direct use of the coordinates-based regression estimation method may have some deviation. It is, therefore, modified into y"=132.979-0.54×latitude (" x_1)"-0.005" 〖×elevation (x〗_2)"-0.9×longitude"(x_3) (r=0.910). In the end, in view of the special terrains of Guizhou, a series of regression estimation models are established to fit regions different in elevation, namely, for regions ≤ 800m in elevation, the model of y"=129.57-0.551" x_1 "-0.006" x_2 "-0.864" x_3(r=0.910) is used；forregions between 800m and 1400m in elevation, the model of y"=133.693-0.514" x_1 "-0.005" x_2 "-0.916" x_3 (r=0.864) is; and for regions above 1400m in elevation, the model of y"=141.454-0.724" x_1 "-0.005" x_2 "-0.945" x_3(r=0.976) is, separately. Results show that the five methods, some ofwhich have been modified, may yield basically similarestimations, which indicates that they are all applicable to estimation of soil temperature regimes in Guizhou. But for regions lacking sufficient soil or air temperature data the method of longitude-latitude-elevation based regression estimation method is more commonly used. The soil temperature regimesin Guizhoucan be sorted into three categories:hyperthermic, thermic and mesic; Guizhou has 80 counties (cities) in the category of thermic temperature regime;Weining and Dafang of Bijie in the category of mesictemperature regime; and four counties in between the two categories, like Shuicheng, Kaiyang and Xishui counties where both mesic and thermictemperature regimes exist, and Luodian where both thermic and hyperthermic temperature regimes do. Besides, when temperature regime is used as diagnostic feature in soil classification in Guizhou, attention should be paid to the fact that the meteorological stations may represent most of the areas under the influence of the atmosphere, but not the maintain areas where variations along vertical zones and microrelieves exist. As two categories of soil temperature regimes exist in Shuicheng, Kaiyang, Xishui and Luodian counties, it is essential to take into account soil forming environmental conditions, especially topographic conditions, in applying the methods to estimation of soil temperature regimes.