The fast Fourier transform (FFT) is commonly used for signal frequency estimation, and the zero-filling method can reduce the quantization error of the amplitude spectrum frequency search interval, but it will increase the computational effort of frequency estimation exponentially. In this paper, a fast and high-precision frequency estimation algorithm based on FFT amplitude-phase combination is proposed. First, the amplitude spectrum and its peak position are determined by using the spectrum sequence of signal sampling and the difference between tail and head samples. Then, the quantization error correction value of the frequency search interval is determined by the spectrum sequence peak position in the amplitude spectrum and the tail sample difference of signal sampling. Therefore, the proposed method utilizes both the position information and the phase information of the peak value of the amplitude spectrum. Simulation results show that the proposed algorithm has lower computational complexity and higher positioning accuracy compared with the FFT algorithm based on amplitude spectrum search only.