Practical Byzantine Fault Tolerance (PBFT) is one of the most important consensus algorithms in distributed systems, which can effectively respond to the threat of malicious nodes. However, PBFT still has shortcomings in terms of arbitrary selection of primary nodes, high communication overhead, and lack of reward and punishment mechanisms. To address this problem, in this paper, we propose a comprehensive scoring-based Practical Byzantine Fault Tolerance consensus algorithm, called CS-PBFT. The algorithm introduces a comprehensive scoring mechanism to evaluate the node’s reliability and overall capability, which is composed of two key metrics: node honor and recommendation scores. Based on this mechanism, the algorithm selects the primary node, slave node, and alternate node to ensure an efficient and secure consensus process. Additionally, the algorithm optimizes the communication process in the Commit and Reply phases of the PBFT consensus protocol, reducing communication latency and improving consensus efficiency. Experimental results show that the improved PBFT algorithm not only improves consensus efficiency but also reduces communication overhead, strengthens fault tolerance against malicious nodes, and demonstrates better scalability.