Emotions are known to influence cognitive performance, particularly working memory (WM) in both its aspects, processing, and maintenance. One explanatory mechanism might be that negative stimuli capture attentional resources, leaving fewer resources for attentional maintenance and processing of information in WM. However, this attentional capture was only investigated using WM tasks in which participants were explicitly asked to process negative items. The aim of this paper was to determine whether explicit processing of emotional stimuli is necessary to impair WM performance, or if their mere presence is enough to capture attention. For this purpose, participants performed a complex span task in which they alternated between memorizing a series of neutral words and processing either emotional images or neutral ones. In Experiment 1, participants were instructed to explicitly process emotional images, whereas in Experiment 2, emotional images were presented without any processing being required. In both experiments, we observed a decrease in memory performance when the images were negative compared to neutral. Whether or not voluntary processing is involved, emotional images seem to capture attentional resources, which in turn leads to a decline in memory performance. These results were discussed in relation to attentional theories and the influence of emotion on the specific mechanisms of WM.