Abstract:By using daily and monthly mean and minimum temperature data from April to September during 1961-2010 from 93 meteorological stations combined with different kinds of cold damage indices in the flue-cured tobacco planting areas of Yunnan province, based on its probability and strength related with weighting exponent, the integrated risk index in spring, summer, early autumn and the whole growth period, was calculated, to in order to analyze the distribution characteristics of low temperature risk index. A regression model related to elevation, latitude and cold damage index was constructed, which can be used for the small grid reckon, revision and division of cold damage index under ArcGIS 9.3. The whole area was divided into five risk levels, this kind of zonation can provide scientific basis for tobacco production to prevent low temperature injury.