Thermal swelling behavior of coal particles during rapid heating was predicted with basic analytical data of the original coal sample. Plastic property dominates thermal swelling of coal particle; therefore, quantitative correlation between elemental analysis and data relating plastic phenomena enabled the prediction of the thermal swelling ratio of coal particles. The data relating plastic phenomena were parameters derived from thermogravimetirc analysis (TGA). By using the TGA data of 55 kinds of coal samples, multiple linear regression analysis was performed to extract several basic analytical parameters for the prediction. As a result, H/C, O/C, and the amount of volatile matter (VM) employed to the prediction. Then, the correlation between thermal swelling ratio during rapid heating and these basic parameters was established. As an additional parameter, the average size of aromatic ring in coal improved the simulation.