The ability to identify the actions performed by others is a vital human skill that depends on the effective allocation of limited brain resources to prioritize attended targets. This experiment aimed to reveal whether top-down preparation plays a facilitatory effect in the identification of basic action categories depicted as static images. We hypothesized that prior information about the action category of the target image through a word cue (climb, jump, run, or walk) would improve identification performance. We employed a within-subject design, incorporating the factors cue validity (valid, invalid, and non-cued) and action category (climb, jump, run, walk), with a ratio of 3:1 between valid and invalid trials. A two-factorial repeated-measures analysis of variance (ANOVA) for the factors cue validity and action category revealed that participants responded faster and more accurately in the valid condition in comparison to the invalid and the non-cued condition. To ensure that we were not cueing any specific motor association, we carried out a control experiment, in which the assignment between response keys and action categories was randomized for each trial. The control experiment confirmed the facilitatory effect of the cue, with faster and more accurate responses in valid in comparison to non-cued trials. These findings highlight the role of category-based attention for efficient action identification. The semantic category of an upcoming action enables the pre-activation of pertinent visual action representations, prioritizing them over irrelevant actions. This, in turn, facilitates the visual processing of specific action instances when they align with the cue.