A critical pattern selection method for full-chip source and mask optimization based on depth-first search is proposed. The growth factor and projection boundary of mask spectra are used to describe the diffraction spectrum characteristics. The critical pattern selection method based on depth-first search is designed to realize the critical pattern selection for full-chip source and mask optimization, so as to obtain all critical pattern groups. Compared to existing methods of the same kind, the proposed method can obtain all critical pattern groups covering all frequency groups. Tachyon Tflex, one of the state-of-the-art commercial computational lithography software from Netherlands ASML company, is used to simulate and verify the proposed method. The results show that the process window obtained by the proposed method is better than that of the Tachyon Tflex method. The proposed method shows better results of selected critical patterns than the reported methods.