Ordinary differential equations (PDB) play an important role in various fields of science, from physics to biology, to model dynamic phenomena. Numerical methods such as Runge-Kutta are often used to solve GDP when an exact solution cannot be obtained. This research focuses on modifying the fifth order Runge-Kutta method to increase accuracy and efficiency in solving certain GDPs. The performance evaluation includes error and computation time analysis to assess the improvements provided by this modification. Experimental results show that the modified method is able to reduce errors significantly without a significant increase in computing time. This study contributes to the development of more accurate numerical methods for solving PDB, especially for applications with high precision requirements. The use of this modified fifth-order Runge-Kutta method can be a better choice for various complex PDB problems that require a higher level of accuracy without significant compromise in performance.