Abstract: The data on ship accidents in coastal waters of China in the period of 2000—2019 are studied with data processing and mining technologies. The investigation focuses on space distribution of accidents at different serious levels for different ship types. The discrete choice model is used to identify the influencing factors for the accidents. A multinominal Logit model is constructed to study the particular effects of each factor on the consequences of an accident. 6 major factors in 17 groups that have significant influences on the consequences of an accident is identified. The factors in accident type group occupy the first place in terms of effect on consequences.