To efficiently utilize the information which can be extracted for target recognition and the character that different polarization channels characterize the same structure signature of a target using different polarization modes to boost recognition performance, a method for full-polarization HRRP recognition based on multitask compressive sensing was proposed. Each singlepolarization HRRP was represented by the atoms adaptively selected from its associated dictionary, and the atoms derived from different dictionaries corresponded to the same index set. Compared with the conventional methods, the proposed method has the significant advantage of exploiting the correlation among single-polarization HRRPs to enhance recognition performance. Experiments were carried out on simulated data, and the results demonstrate the efficiency of the proposed method.