The sandwich priming paradigm, involving presenting the target as the initial prime in a masked priming sequence (e.g., white-whete-WHITE), has been shown to increase the size of masked orthographic priming effects. It also appears to increase the size of masked morphological priming effects (e.g., walk-walker-WALK > walker-WALK), although not pseudo-morphological priming effects (e.g., corn-corner-CORN = corner-CORN) (Lupker & Spinelli, Journal of Experimental Psychology: Learning, Memory, and Cognition, 49 (11), 1861–1880, 2023). The empirical question addressed in the present research was whether masked semantic priming effects (e.g., black-WHITE) would also increase when using sandwich priming (e.g., white-black-WHITE). Such a result could imply that semantic and orthographic priming have similar sources as well as that Lupker and Spinelli’s presumed increase in morphological priming may have been due to the prime (e.g., walker) and target (e.g., WALK) being semantically related. The results of two experiments, one using conventional masked priming and one using sandwich priming, clearly indicated that sandwich priming does not increase the size of masked semantic priming effects, consistent with the idea that the presumed impact of sandwich priming on morphological priming effects is truly a morphologically based phenomenon.