There is an urgent demand for understanding the maximum temperature resistance of reinforced concrete columns in buildings, as columns determine a building’s resistance to progressive collapse. When a fire occurs, columns are subjected not only to loading but also to fire, making their behavior very complex. Although there are a considerable number of experimental investigations on the fire resistance of reinforced concrete columns under axial force, an accurate and reliable model for predicting fire resistance is still needed to reduce time and cost associated with experiments. In this research, two machine learning (ML) techniques, the decision tree model (DTM) and the support vector machine model (SVMM), have been utilized to estimate the maximum temperature resistance of columns subjected to axial force. A total of 300 experimental test results have been collected to train and test the ML models, considering the effects of 12 input variables. The outcomes of the predictions revealed that among the proposed ML models, DTM exhibited excellent performance in predicting the maximum temperature resistance of columns in both the train and test sets with high accuracy and reliability. The R values from DTM were more than 0.96, and the RMSE values were below 10% of the average maximum temperature. From the sensitivity analysis, column length served as the most important parameter influencing the maximum temperature resistance of columns, as indicated by DTM, whereas the boundary condition was highlighted by SVMM.