Considerable controversy has centered on the origin of the observed two peak structure in photoemission from Ce compounds. We report results of theoretical studies of the Ce pnictides using a model based on competing screening mechanisms. It is assumed that the peak near EF arises from complete screening of the 4f hole via repopulation of the f orbital on the hole site so its contribution to the spectrum follows the ground state density of states. The second peak 2-3 eV lower in energy is viewed as being due to a poorly screened f-hole. This screening process is studied by means of self-consistent LAPW supercell calculations. In CeSb we find a peak separation of 2.71 eV compared to an experimental value of 2.5 eV, whereas in CeP we find a separation of 2.47 eV compared to 2.4 eV experiment. The poorly screened peak results primarily from d-orbital screening of the hole.