In view of the problem of intricate overlap and difficult separation in midcourse group-target echo by narrowband radar, a novel algorithm based on matched space transform was proposed to extract its micro-Doppler features.On the basis of modeling of group-target, period of each sub-target was obtained using autocorrelation method first.Then with the characteristic of micro-Doppler curve along time variable, the corresponding matched space was structured.Finally, the parameters of each scattering point were estimated utilizing energy peak searching and CLEAN algorithm.Thus the echo was reconstructed.Simulation results indicate that the proposed method can achieve favorable identification despite of background noise and signal aliasing.