Fire is categorized under accidental loads in Eurocodes. Accidental fire loads are almost unpredictable loads that depend on time, the fire compartment, live loads, and the type of such loads. In case of complicated structural systems, some aspects—such as the buckling of columns, the effects of elevated temperatures on global structure, the redistribution of loads and internal forces, and the unloading process—can be investigated only by applying advanced calculation methods based on the Eurocode classification, which apply the coupled finite element fields to analyse the structure. The solution process of complicated systems can be time-consuming, costly, and—sometimes—impossible. The abovementioned methods are complicated and time-consuming for modelling the accidental fire loading process as different scenarios in combination with other loads. The motivation of this research study is to simplify the modelling of the fire loading processes. By combining the simulation procedure of Levels 2 and 3, based on Eurocode, a FEM model of analysis for mathematizing the fire loading process is developed. Based on the results of the two-dimensional thermal FEM equation, the developed method obtains the reduced time-dependent stiffness (integrated over the cross-section area) of warm members as a function of time. The variable stiffness of warm members will be applied to mathematize the global structural behaviour for calculating the distribution of internal forces, considering geometrical nonlinearity. The developed method is able to apply different fire scenarios and combine them with other service loads to obtain global structural behaviour and distribution of internal forces. To achieve this, a FEM code has been developed in Microsoft Visual C# (C sharp), and the results are verified with those of the ANSYS Workbench software. To investigate the fire resistance of steel structures, the method that has been developed will give convincing results at lower temperatures. The developed method considers a cold analysis under an assumption of variable stiffness for the warm members and, so, is suitable to just determine an estimation. The developed method attempts to decrease the cost of analysis and give a rough estimation of the structural response. The results of the developed method can be trusted to obtain an estimation of the global behaviour of structure. The results depend directly on stiffness curves. The simplified method that has been developed is not suitable to determine exact and accurate structural behaviour. Material nonlinearity, crashes in concrete members, 3D thermal stresses and strains in cross-sections, buckling, and shear deformations are some of the parameters not considered by the method that has been developed.