In recent years, owing to an increase in electrical energy consumption, the quality and safety issues of power systems have drawn considerable attention. The frequency of a power system is an important indicator of its stability. The system frequency will drop when large generator sets are tripped. Therefore, it is important to measure the fundamental frequency of a power system accurately. The power system frequency can be estimated through various methods in time or frequency domains. Among these methods, curve fitting is a time domain approach that can be used to identify the parameters of a model using the input information obtained by utilizing a nonlinear regression method to fit the input curve. The physical phenomenon of a power system is described by a mathematical model. The curve fitting approach is applied to find the parameters such that the model is closer to the measured signal. These parameters are used to obtain the fundamental frequency of the power system. In this paper, the genetic algorithm (GA) is compared with the conventional regression analysis (RA) method for identifying the parameters of the model. The performance of curve fitting using different mathematical models on various power system events is discussed.