In order to improve the intelligent level of evidence collection and responsibility determination of traffic accidents, this article proposed a vehicle information management and accident forensics system based on blockchain and consensus mechanism. Vehicle information was managed in a permissioned blockchain framework, and the communication between the Electronic Control Unit (ECU) of the vehicle and the Road Side Unit (RSU) was used to ensure the legitimacy and integrity of vehicle data. In the proposed traffic accident digital forensics scheme, the data from involved vehicles was automatically managed by the RSU, and a consensus on data reliability was achieved through the practical Byzantine Fault Tolerance (pBFT) protocol. The experimental results show that the proposed blockchain scheme is able to meet the real-time requirements of transportation applications, and the block generation delay and Q-A verification delay in the experimental scenario do not exceed 18.13 ms and 1.55 ms, respectively. The qualitative results show that the proposed framework is able to resist known security attacks, ensure data reliability, and can help law enforcement achieve fair traffic accident liability determination.