In this paper, we present a discrete fractional evolutionary transform (DFrET) for the time-frequency (TF) representation of non-stationary, wide band signals. The time-varying kernel of this transform is used to calculate the evolutionary spectrum. The DFrET kernel are obtained from the coefficients of a discrete fractional Gabor expansion. The proposed DFrET provides a tool for high-resolution representation of multicomponent signals with linear instantaneous frequencies. Performance of the proposed algorithm is illustrated by means of simulations and compared with existing TF methods.