Precursors to P′P′ (PKPPKP), first interpreted as sub-surface reflections by Gutenberg in 1960 and studied in several later papers by other authors, precede the P′P′ phase by up to 200 sec. This phase, designated P′dP′ where d is the depth of reflection, has unique potential for giving new details of upper-mantle structure. However, as with any newly discovered seismic phase, the uniqueness of its interpretation must be well established. Asymmetric P′P′ phases reflecting from surface or near-surface dipping interfaces pose a challenge to this uniqueness because of their maximum-time nature. Simplified estimates of the amplitudes of asymmetric P′P′ rays are made, including consideration of the relative amplitudes of core phases and the finiteness of the reflecting surfaces of dipping interfaces. These estimates lead to the conclusion that the reading of asymmetric P′P′ at a single station is likely only in the 0- to 30-sec range before the main symmetric P′P′ phase. However, if array beam-forming is used, this range is reduced to 0 to 10 sec. The data indicate that both P′dP′ and asymmetric P′P′ are present at up to 30 sec lead time and array beam-forming is needed to differentiate between the two. A further effect of the maximum-time nature of P′P′ is that, in practice, the geographic location of the reflection point can be determined to within only a few degrees.