Abstract: Based on the Designer’s Simulation Toolkit (DeST), a method for predicting the dynamic embankment heating load in cold regions are proposed. A model composed of a series of 'micro elements' is first established, and the meteorological, shielding, and illumination models are used to simulate the climatic environment, topographic conditions, and solar radiation of the embankment, respectively. The state space method in the Building Analysis Simulation (BAS) module is then used to determine the initial temperature. The accuracy of the calculation is increased by setting the similarity threshold between the predicted value and the model training data. The heating load is identified by setting the target temperature. As a case study, the method is applied to the DK221+150 section of the Harbin-Qiqihar high speed railway. The results show consistency between the calculated and measured temperatures. The maximum and average heating loads are 945 W/m and 335 W/m, respectively. The method reflect the effects of both slope exposure to solar radiation and the lag in heat transfer inside the embankment. The data for typical meteorological years of 23 cities positioned in cold and severely cold climate zones in China are selected for a regional statistical analysis of the heating load of a high-speed railway embankment. The results show a maximum heating load of 531~ 1,338 W/m and a linear positive correlation between heating load and latitude. The results of this study can guide the evaluation of embankment frost state and the design of heating schemes in cold regions.