Stochastic climate generators are used in many studies such as application of hydrologic, environmental management and assessment of agriculture risk. These studies require for assessment of risk to long term series from meteorological data. Considering of climate data limitation at large area in Iran and short term of data, it is necessary using of climate generator and evaluation of precision and accuracy. Therefore in this study have been evaluated the efficiency of three generators namely CLIGEN, ClimGem and LARS-WG in sangeneh and Zidasht station with different climate condition. Statistical test of t (t paired) have been used to compare the differences between observed and production weather data such as yearly and monthly precipitation amount, yearly number of wet day, yearly average of max. and min. temperature. The obtained results show that CLIGEN generator have the better efficiency than two others generator in two stations and five considered variables. ClimGen haven't had the good efficiency in two stations. Also, LARS-WG generator have had the good efficiency in Zidasht station, but it's efficiency have had the less efficiency for product of temperature variables in Snganeh station. Generally, the obtained results show that the efficiency of these generators is better in mild climate than arid and semi-arid climate.