In addressing the challenge of uneven passenger flow distribution within urban rail transit networks, this study introduces an advanced multi-path passenger flow assignment model grounded in regret theory. The model meticulously constructs a multi-routing space-time topological network and incorporates regret theory to accurately depict passengers’ decision-making processes. By taking into account key factors such as perceived travel time, the number of transfers, and train congestion levels, the model aims to allocate passenger flow based on the principle of minimizing passengers’ anticipated regret. This innovative approach not only enhances the balance between supply and demand but also significantly improves operational efficiency. The findings of this study provide robust theoretical support and practical solutions for optimizing urban rail transit network management, offering a novel perspective on achieving more efficient and passenger-friendly transit operations.