Display omitted •Adopted SR to mine the fatigue mechanism based on test data.•Developed a fatigue equation to determine the fatigue life.•Demonstrated the accuracy and reliability of the developed fatigue equation.•Data-driven model for determination of fatigue equation. The fatigue life is one of the essential design indexes of the cement-stabilized cold recycled mixtures (CSCRM) in China. However, it is challenging to determine the fatigue equation of the CSCRM. In this study, the symbolic regression method is adopted to automatically identify a closed-form equation by the data-driven model to establish the fatigue prediction equation with a unified and concise mathematical structure. The developed method was illustrated and compared with the test results. The results show that the developed equation has a satisfactory prediction effect and provides a good and helpful tool to determine the fatigue life for CSCRMs.