Abstract: Cold chain multimodal transport combines the particularity of cold chain transportation and the complexity of multimodal transport. Compared with general transportation problems, its path optimization model is more complex, and the requirements for algorithms are also higher. A review study was carried out on the path optimization problem of cold chain multimodal transport, focusing on models and algorithms. First, a knowledge graph visual analysis was performed on 369 Chinese and English-language literatures screened by the systematic literature review method. Then, a key analysis was conducted on 40 of these literatures that simultaneously contained the construction of cold chain multimodal transport path optimization models and algorithm solutions. The selection of optimization objectives and cost calculation, along with model-solving algorithms were summarized. It is found that in the single-objective optimization research of cold chain multimodal transport, the optimization of the total transportation cost is the main focus. In multi-objective optimization research, the objectives mainly include minimizing transportation cost, minimizing transportation time, maximizing customer satisfaction, and minimizing carbon emissions. Although multi-objective optimization is more in line with the actual situation, it has problems such as complex models and high computational complexity. Therefore, most literatures still choose single-objective optimization. Regarding the selection of calculation methods for refrigeration cost, cargo-damage cost, and carbon-emission cost, most studies calculate with fixed cost coefficients and seldom consider the impact of different transportation conditions on cost calculation. Solution algorithms mainly fall into two categories: exact algorithms and non-exact algorithms. Exact algorithms are suitable for small-scale optimization problems, while machine learning algorithms among non-exact algorithms are often combined with other algorithms and demonstrate greater advantages in solving large-scale practical problems. Finally, future development directions are prospected from aspects of model construction, objective selection, computation, and algorithm implementation.