In order to reduce the accident rate and improve the traffic efficiency， the influence of meteorological， traffic， time， and bridge structure data on the traffic accident rate of the estuarine long-span bridge was studied， based on the multi-source real-time data provided by the health monitoring system of Jiashao Bridge. The time matched sampling method and random sampling method were used to extract the accident data within four years of operation， and a conditional logistic regression model was established. The results indicate that the traffic flow and humidity are significant positive factors for the probability of traffic accidents of estuarine long-span bridges while the average traffic speed is a significant negative factor for the probability of traffic accidents. During peak hours， the possibility of accidents increase significantly. According to the existing data， the deflection change of the bridge， structural vibration， and the proportion of trucks in the traffic flow have no significant influence on accident rate.