Time-frequency transform（TFT） resembles an ideal complex band-pass filter. Based on this property， TFT can be used to measure the synchronous complex phasor by filtering the fundamental frequency band of the signal， without being interfered by other frequency bands. The width of frequency band and the length of data window required for the calculation are both decoupled from the centre frequency and can be flexibly controlled by the scale parameter. The real part and imaginary part of TFT have the consistent frequency-magnitude characteristics. Therefore TFT can adapt to the offset of the real system frequency. A novel algorithm for the synchronous phasor measurement based on TFT is proposed. By constructing the new time-frequency atom-function and selecting the optimal scale parameter， TFT can be used to preciously measure the fundamental magnitude， frequency and phase of the signal with the fixed interval sampling while overcoming the effects caused by the large-scale offset of frequency， harmonics， noises and out-of-band components. It is worth noting that the tracking and compensation due to frequency offset are not required in this process. Moreover， TFT has ideal dynamic characteristics with the fast response speed for the step change of magnitude， frequency and phase. Finally， the correctness and validity of the proposed algorithm are well verified by the MATLAB and EMTDC/PSCAD simulation results.