Emotional information captures attention due to privileged processing. Consequently, performance in cognitive tasks declines (i.e. emotional distraction, ED). Therefore, shielding current goals from ED is essential for adaptive goal-directed- behaviour. It has been shown that ED is reduced when participants recruit cognitive control before or after the presentation of an emotional negative distractor. Following up on this, we asked first, whether cognitive control of ED is negative-valence-specific or valence-general. A valence-general-account predicts that control shields against distracting influence of emotion, irrespective of the specific valence. In contrast, a negative-valence-specific-account predicts that control interacts with the valence and ED is reduced for negative stimuli only. Second, we asked whether this effect of ED differs between control modes operating on different time scales (i.e. proactively or reactively). To test this, we manipulated emotional distractor valence (positive/high-arousal; negative/high-arousal; neutral/low-arousal) and assessed how control interacts with ED. Results showed that ED was reduced for negative and positive valent stimuli when control was triggered before (i.e. proactive control, nExp1 = 141, between-subject-design) and after (reactive control, nExp2 = 37, within-subject-design) the emotional stimuli. Accordingly, control blocks off high-arousing emotional distractors from interfering with goal-directed-actions, irrespective of their valence (i.e. valence-general-account) and for both, proactive and reactive control modes. [ABSTRACT FROM AUTHOR]