The study of nonconscious priming is rooted in a long research tradition in experimental psychology and plays an important role for a range of topics, including visual recognition, emotion, decision making, and memory. Prime stimuli can be transiently suppressed from awareness by using a variety of psychophysical paradigms. The aim is to understand which stimulus features can be processed nonconsciously and influence behavior toward subsequently presented probe stimuli. Here, we tested the notion that continuous flash suppression (CFS), a relatively new method of interocular suppression, selectively disrupts stimulus identification mediated by the ventral “vision-for-perception” pathway, while preserving action-relevant stimulus features processed by the dorsal “vision-for-action” pathway. Given the far-reaching implications of this notion for the influential two visual systems hypothesis, and visual cognition in general, we investigated its empirical basis in a series of seven masked priming experiments using CFS. We did not find evidence for nonconscious priming of object categorization by action-relevant features. Based on these results, we recommend skepticism about the notion that the processing of action-relevant features under CFS is selectively preserved in the “vision-for-action” pathway. Second, we conclude that CFS experiments are less informative than approaches using visible stimuli, when the aim is to gather data in relation to the two visual systems hypothesis. Third, we propose that future nonconscious priming studies should carefully consider the position of suppression paradigms within a functional hierarchy of unconscious processing, thus constraining hypothesis generation to effects that are plausible given the employed methodology.