Recent research indicates that emotional faces affect motor control only when task-relevant. However, these studies utilized a single-face presentation, which does not accurately mirror real-life situations wherein we frequently engage with multiple individuals simultaneously. To overcome this limitation, we gave 40 participants two versions of a novel Flanker-Go/No-go task, where we presented three-face stimuli with a central target and two task-irrelevant flankers that could be congruent or incongruent with the target for valence and gender. In the Emotional Discrimination Task (EDT), participants had to respond to fearful or happy targets and refrain from moving with neutral ones. In the Gender Discrimination Task (GDT), the same images were shown, but participants had to respond according to the target's gender. In line with previous studies, we found an effect of valence only in EDT, where fearful targets increased reaction times and omission error rates compared to happy faces. Notably, the flanker effect, i.e., slower and less accurate responses in incongruent than congruent conditions, was not found. This likely stems from the higher perceptual complexity of faces than that of stimuli traditionally used in the Eriksen Flanker task (letters or signs), leading to a capacity limit in face feature processing.