Abstract: In order to estimate the risk degree of agricultural machinery accidents accurately, explore the causes and laws of agricultural machinery accidents, and prevent the occurrence of various accidents in the process of agricultural machinery products effectively, this study constructs a fuzzy fault tree model by combining the fault tree analysis method with the triangular fuzzy number and analyzes the factors which cause agricultural machinery accidents. The results showed that 23 known and potential causes of agricultural machinery accidents were deduced, and 57 minimum cut sets and 3 minimum path sets were solved through fault tree analysis. The probabilities of 23 basic events were fuzzed by triangular fuzzy number theory, and the fuzzy probability value of each basic event was obtained. Sorting the fuzzy importance of 23 basic events clarified the degree of influence of each basic event on the top event. Machine without safety protection device（X₁）, unprotected individuals（X₂）, long working hours(X₁₆), flawed control device(X₁₅), speeding and overloading（X₇） were ranked first. It shows that these factors greatly impact the occurrence of agricultural machinery accidents. When formulating accident prevention measures, priority should be given to formulating preventive countermeasures to scientifically and efficiently prevent agricultural machinery accidents.