To improve the accuracy of harmonic analysis, the signal-to-noise ratio loss caused by signal windowing and the quantization error caused by AD conversion is analyzed, and the principle of improving signal-to-noise ratio by oversampling technology is expounded. A harmonic analysis method based on high-rate oversampling and windowed interpolation FFT is proposed. This method makes full use of the potential of an AD converter to take samples quickly, and at the same time, it avoids the surge of sampling data caused by high oversampling through mean filtering. The influence of the proposed method on the amplitude and phase of harmonic components in the signal is studied in detail, and a concise and practical correction method of harmonic amplitude and phase is given. Simulation results show that the proposed method can improve the anti-noise ability of windowed interpolation FFT and the accuracy of harmonic analysis without increasing system cost.