By taking into consideration the uncertainty of the weather condition and system operating status， the model for certainty between load and outage rate is converted into a fuzzy one as wind or ice is certain to cause load influence on the transmission components outage rate. Then the fuzzy if-then rules of three input variables and fuzzy inference system are built to model the condition-based outage rate， and defuzzification is applied to obtain an exact non-fuzzy（crisp） value from a fuzzy set to evaluate power system reliability in a particular adverse weather condition. According to the weather condition based line outage rate model， multi-area reliability levels are solved and analyzed. Suppose 5 lie lines in RTS-79 are influenced in normal weather， adverse weather and extremely adverse weather conditions respectively. The line outage rates and reliability levels are calculated under different weather conditions to analyze the influence of weather conditions on the multi-area system reliability.