Abstract: With the large-scale integration of new energy, the inertia of the power system decreases, and the frequency fluctuations in the power system become severe, increasing the difficulty of accurate frequency measurement. The frequency measurement accuracy directly affects the effectiveness of various frequency control applications in the power system. Therefore, this paper proposes a novel, accurate frequency measurement method from the perspective of circular motion. This method establishes the circular motion representation model of the rotating phasor of the sinusoidal voltage signal. Then, based on the trajectory motion equations of the uniform and non-uniform circular motions, the functional relationship between the rotating phasor and its derivative and the system frequency is derived. Furthermore, using quadratic polynomials to fit the dynamic amplitude and phase angle of voltage signals, a method based on the least squares method for solving rotating phasors and their derivatives is put forward. In addition, an iterative optimization method based on a frequency grid is proposed to eliminate the impact of frequency offset on the algorithm. Simulation and hardware test results show that the proposed algorithm can accurately measure the fundamental frequency under frequency offset, linear, and nonlinear changes.