In the present study we explored the impact of task-irrelevant emotive picture stimuli on visuo-spatial vigilance performance, self-reported state and memory. Ninety-five participants (62 women, 33 men) completed the experiment in which task-irrelevant emotive picture stimuli were embedded in the vigilance task. Four experimental groups were tested by combing two levels of valence, positive versus negative, and two levels of arousal, arousing versus neutral or non-arousing, for the task-irrelevant picture stimuli. The vigil was organised so that baseline performance, the initial impact of the images, and any continual carry-over effects of the images on performance could be measured. In addition to performance on the vigil, subjective state was measured using a self-report questionnaire designed to examine energetic and tense arousal as well as task-related and task-unrelated thoughts. A post-task freerecall test was also employed, asking participants to recall as many of the picture stimuli as they could. Results showed a significant arousal by period linear trend interaction, in which the performance decrement of the groups exposed to the arousing picture stimuli was attenuated in comparison those exposed to the nonarousing stimuli. Further the relationship between self-reported energetic arousal and performance differed for the arousing and non-arousing picture groups. Post-task energetic arousal significantly predicted the performance decrement (linear slope) for the arousing picture group, but not for the non-arousing picture-group. The arousing pictures were also recalled at a higher rate than non-arousing pictures, irrespective of valence. These results provide support for the perspective that the arousal quality of picture stimuli matters more for performance than valence, and that arousing pictures while possibly disruptive when presented concurrently with the vigilance task, may result in improved performance later due to an increase in energetic or cortical arousal. This finding fits with previous research suggesting that arousing agents are more potent when their possible distracting effects on task performance are no longer competing for cognitive resources.