Bicycle accidents have increasingly caused casualties and property damage. Many countries have therefore started to pay more attention to designing the space of bicycle paths. However, few studies have focused on the design of bicycle continuity between each bicycle path. In this study, the concept of using both spatial crash probability (P) and crash severity index (CSI) is introduced to address the bicycle safety issue on bridges in Central Business Districts (CBDs). Bicycle paths on bridges usually face design difficulties due to limited space, different altitudes and the amount of traffic volume. Therefore, a systematic process should be established to evaluate and implement cycling space on bridges. The characteristic of bicycle-motorized vehicle collisions (BMV), traffic engineering, road environment and driving behaviour are analysed through spatial negative binomial modelling (NB). It is the aim to give recommendations to city planners and governments concerning the enhancement of bicycle safety and riding continuity on bridges.

bridge; continuity; bicycle paths design; spatial negative binomial modelling (NB); crash severity index (CSI); bicycle-motorized vehicle collisions (BMV); and geographic information system (GIS)