Abstract: An algorithm using FFT to calculate Wignev-Ville Distribution (WVD) is introduced. In view of the complexity of time frequency distribution numerical calculation of Wigner-Ville quadratic function, and the large amount of calculation by the algorithm, a sampling algorithm is put forward, by which the length of data that can be handled gets increased in the condition of not changing the sampling frequency. Or, when dealing with the same data length, the calculation time is shortened greatly. The algorithm is applied to analyze the time frequency distributions of a simulated time varying sinusoidal signal and of the fast changing noise signal produced by closing a steel window, and satisfactory results are obtained.