Abstract: In order to study the relationship between meteorological factors of different ecotopes and forest fire during 19822008, the annual mean wind speed, daily temperature range and wet coefficient were selected as indicators of meteorological elements, forest suffer rate as indicator of forest fire based on dividing Yunnan Province into five ecotopes. The results showed that: 1) meteorological factor of each ecotope changed obviously, annual mean wet coefficient and wind speed changed relatively synchronous, showing a whole significantly down trend, but the annual mean daily temperature range changed asynchronous. 2) Annual forest suffer rate of each ecotope presented a significant downward trend and showed inverse curve from 1982 to 2008, and the changing pattern of annual forest suffer rate varied with time series and showed a significant “S” function. 3) Different meteorological factors of different ecotopes had distinct influence on forest suffer rate, which was affected by annual mean wind speed and the annual mean daily temperature range more heavily than by the annual mean wet coefficient. Using logistic regression model can well explain the relationship between meteorological factor and forest suffer rate. 4) Due to the difference of vegetation type, topography and other factors, forest fire of each ecotope were affected differently by meteorological factors. In the area of ecotopeⅠ,which was covered by nonflammable fuels (evergreen broadleaved forest), lower elevation, and dense population , meteorological factors had limited impact on forest fires. While in the area of ecotopeⅤ,which was covered by flammable fuels (coniferous forest), higher elevation and sparse population, meteorological factors had larger impact on forest fires. Forest fire was the result of various influencing factors together. Therefore, in the forest fire prevention division, dividing by ecotope is more reasonable, which can provide scientific basis for decision making and management of fire prevention for fire department.