The aim of this paper was to study whether real angry faces do capture attention to the extent of overcoming the inhibition of return (IOR) effect and whether the anxiety level of participants modulates this effect by stressing biases toward threatening stimuli. With this purpose, participants categorized the emotional valence of face targets in a standard spatial cueing procedure suitable to measure IOR. In Experiment 1, participants were selected according to their high vs. low-trait anxiety, whereas in Experiment 2 participants were induced a positive vs. anxiety mood state. The typical IOR effect was observed with neutral and happy face targets, which disappeared with angry face targets. Similar results were observed for all anxiety groups and in both experiments. The results indicate that IOR is overridden when the target is a biologically relevant angry face, as highly relevant targets should suffer less from habituation to attentional capture regardless of anxiety. We suggest that these data show that attentional capture is less likely to habituate for threatening information, so that no cost is measured in detecting new threatening information appearing at recently cued locations.
