A traffic matrix estimation approach for optical transportation networks was proposed.Compressive sensing theory was used to study traffic matrix estimation in optical transportation networks.According to the sparse representation of signals,traffic matrix was processed in the sparse way.Matrix transform theory was exploited to present a new network tomography model for optical transportation networks.This model can overcome the illposed nature of the existing network tomography.Conex optimization was used to attain the estimation equation about traffic matrix.The detailed estimation algorithm is presented.The accurate estimation about traffic matrix for optical transportation networks was obtained.The data from the real network was used to perform the simulation.Simulation results show that the proposed method is effective and feasible.