Abstract: To explore the temporal and spatial characteristics of accident causes and the driving principles of bicycle accidents, and to improve urban roadway traffic safety management, by using geographically and temporally cluster analysis and geographically and temporally weighted regression models, bicycle accident data from 2007—2019 in North Carolina were selected to explore the geographically and temporally distribution patterns of factors influencing the severity of accidents, the potentially unobserved heterogeneity, and the trend of the influence of each factor on the severity of accidents under different spatial and temporal conditions, aiming at targeting the reduction of both local and global traffic accident probability. Results show that in the kernel density estimation considering seasonal characteristics, winter accidents show a high incidence; in the spatial cluster analysis exploring different seasons, accident severity is non-stationary in time and space, forming the high/low clustered and the hotspot clustering distribution in summer and winter in six regions including Wake and Durham. Under the condition of season as a time variable, the eighteen of the 24 explanatory variables, including alcohol or drug use by both parties to the accident, high road speed limit, hit-run, rider age over 55, ambulance call, and curved road, had significant positive effects on accident severity, and the geographically and temporally weighted regression model handled the spatial-temporal heterogeneity of the data well. Due to the combined effects of multiple influencing factors in diverse spatiotemporal contexts, aggregation exhibits differentiated distribution patterns. Based on the above findings, this paper investigates the underlying formation mechanisms and formulates targeted policy suggestions to enhance the traffic safety level of vulnerable road groups. The study provides a reasonable theoretical basis for practitioners related to traffic safety and management, and plays a role in improving the travel safety of bicyclists.