The.selection and quantization of landslide factors are important in the classification evaluation of landslide hazard.The landslide evaluation factors are composed of common characteristics factors and non-common characteristics factors.The quantificational indexes of common characteristics factors have consistency in evaluation which could not reflect the difference in the individual landslide evaluation system.Based on the landslide monitoring and early warning data in Wanzhou,Chongqing,the four non-common characteristics factors,i.e.stratum ob-.liquity,gradient,relative elevation and ratio of length and breadth of landslide,were selected in the classification evaluation of landslide hazard.The quantificational evaluation indexes of factors were developed by use of Lognormal distribution model.According to the five ranks evaluation indexes,the landslide data were tested.The test results indicate that the landslides proportion with over-medium hazard is 95.6% and that with high hazard and very high hazard is 76%.The results show that the evaluation factors are reliable and easy to be measured and obtained and the method could be applied to classification evaluation of landslide hazard of regional landslide disasters point and landslide hazard regionalization.