Abstract: [Objective] Developing forest fire insurance pure rates (P) model which adapt to China's current forest insurance system and making analysis on forest fire insurance rates distribution by regional in China can provide the method and reference for rating the forest fire insurance premium rate scientifically and reasonably.[Method] This paper developed a forest fire insurance pure premium rate model basing on Holecy's forest fire insurance premium rate model(Holecy et al., 2003; 2006). The Expected pure premium rate (EP)is mean of forest fire occurrence probability and the Risk pure premium rate(RP)is tolerable error of EP. The model involved the insurance rate fire occurrence probability and other factors. It replaced the Holecy model's insured area variable and changed the measurement units from the actual value to thousand points(‰).To check the model, it make an analysis on China's forest insurance pure premium rate under different insurance rates by regions in 2014 and compare the estimation results(P_βM) and the actual rate(P^*_βM) of 19 forest insurance modeling regions by paired sample T test.[Result] It indicated that the EP of most regions in China were below 1.00‰ but regional differences was significant. The P shall increase with the decreasing of the insured rate. The P of 19 forest insurance modeling regions is 0.05% to 7.20% as the current insured rate. There was no significant difference between the actual pure premium rate and estimated pure premium rate according to the paired samples T test.[Conclusion] The model considered both the identified and uncertain factors and fitted current forest insurance system in China. The estimated result was accurate. It recommended that the data should be adjusted dynamically according to the time and spatial variation when using the model. This model can be use to determine the forest fire insurance gross premium rate.

Key words: forest insurance, pure premium rate, rate making, forest fire