Abstract: To effectively reduce the risks associated with hazardous materials road transportation, this paper addresses the point-to-point hazardous materials transportation issue. It considers path diversity within the k-shortest path problem and proposes a path diversity calculation method combining "grid" and "buffer zone." The paper establishes a dual-objective optimization model that takes transportation risks and path diversity with vehicle capacity constraints into account, and designs a solution based on the Non-dominated Sorting Genetic Algorithm II （NSGA-II）. To verify the algorithm's efficiency, an example with 10 nodes is used. First, the k-shortest algorithm is applied to obtain 15 initial paths that meet the minimal risk requirement of the road network. Then, based on different transportation volumes, various path combinations are selected to compare risk values and diversity under different buffer zone radii. The experimental results show that, when the k-value is limited to 10, changes in the buffer zone radius have a significant impact on path diversity. At the same time, the hazardous materials transportation risk value fluctuates by less than 5% as the buffer zone radius changes. The k-shortest path diversity optimization method can consistently generate transportation route clusters, thus better balancing hazardous materials transportation risks and path diversity. The example calculations demonstrate that, when the k-value is fixed, variations in the buffer zone radius have little significant impact on total risk. However, the buffer zone radius is a key factor in determining route diversity. When the radius is large, transportation routes are considered overlapping, and only when the buffer zone radius is small can the model calculate more diverse routes. Therefore, the buffer zone radius needs to be precisely calibrated according to actual conditions and should not be arbitrarily selected.