本文研究對影響交通事故嚴重程度之決定因素，國外學者提出地區因素作為危險分類之一，則是在美國及澳洲地區行之有年。不同的地區其氣候、道路狀況、人口密度、交通流量與生活水準都不一樣。美國地區性因素便是一項非常重要的分類因素，預期損失因地區性的差異程度比從人及從車因素更大。藉由觀察其他因素影響交通事故嚴重程度，重新打散現行汽車保險依據被保險汽車之從車因素及被保險人之從人因素之風險分類概念，進而探討道路種類與汽車任意責任險風險分類之間的關聯性，可否作為汽車責任險商品費率計算的新風險分類因子，俾使商品費率計價制度更趨公平合理。藉由資料樣本觀察、因子之間獨立性檢定、建立假說及回歸模型進行實證研究，而本研究之結果支持道路種類與交通事故嚴重程度具有關聯性，可作為提供產險公司未來納入汽車任意責任險商品費率計價的新風險分類因子之參考。 This paper studies the impact determinants of traffic accident severity, foreign scholars list the regional factor as one of the risk classifications, and has been adopted in U.S. and Australia for years. The climate, road conditions, population density, volume of traffic and the standard of living all varied in different regions. Regional factor is a very important classification factor in the U.S., the level of difference in expected loss from regional factor is much more greater than driver or vehicle factor. By observing other factors that affect the severity of accidents, breaking the concept of driver or vehicle factor which is followed by current auto insurance when considering the insured risk classification, and to explore the correlation between road type and auto liability insurance, whether it could serve as a new risk classification factor on calculating the rates of auto liability insurance products, and to work toward the goal of a more reasonable and fair product pricing. Through data sample observation, independence test between factors, establishing regression models and hypotheses empirical research, the results of this study support the existence of correlation between road type and severity of accidents, and would be useful for providing auto insurance companies the reference on new risk factors of pricing on auto liability insurance as further consideration in the future.