Previous findings suggest that planning an action "backward-primes" perceptual dimension related to this action: planning a grasp facilitates the processing of visual size information, while planning a reach facilitates the processing of location information. Here we show that dimensional priming of perception through action occurs even in the absence of active action planning. Subjects watched video clips showing a grasping or reaching action before detecting size- or location-defined deviants in visual stimulus sequences. Size deviants were detected faster after seeing a grasp and location deviants were detected faster after seeing a reach. This supports the assumption that perceptual codes and action plans share a common representational medium, and that "attention to action" controls "attention to stimuli". (Contains 1 table, 2 figures, and 1 note.)