Schematic threatening, friendly, and neutral faces were used to test the hy
pothesis that humans preferentially orient their attention toward threat. U
sing a visual search paradigm, participants searched for discrepant faces i
n matrices of otherwise identical faces. Across 5 experiments, results cons
istently showed faster and more accurate detection of threatening than frie
ndly targets. The threat advantage was obvious regardless of whether the co
nditions favored parallel or serial search (i.e., involved neutral or emoti
onal distractors), and it was valid for inverted faces. Threatening angry f
aces were more quickly and accurately detected than were other negative fac
es (sad or "scheming"), which suggests that the threat advantage can be att
ributed to threat rather than to the negative valence or the uniqueness of
the target display.