The behavioral urgency hypothesis suggests that stimuli signaling potential danger will receive attentional priority. However, results from the gaze cueing paradigm have failed to consistently show that emotional expression modulates gaze following. One possible explanation for these null results is that participants are repeatedly exposed to the same emotional expressions during the typical gaze cueing procedure. We employed a relatively novel gaze cueing method in which participants were presented with 2 unique (or "rare") trials during an experimental block. Specifically, either 2 fearful face trials appeared within a block of happy faces or 2 happy face trials appeared within a block of fearful faces. Results showed that when participants were repeatedly exposed to the same emotional expression, gaze cueing was independent of face type. However, when the emotional expression was a rare event, significantly larger cueing occurred for fearful than for happy faces. These results support the behavioral urgency hypothesis and show that emotional expression does indeed modulate gaze following.