Items that are distinctive with respect to their context tend to be recalled better than nondistinctive items, a finding known as the von Restorff effect. The goal of this study was to elucidate the role of novelty in this effect. In two experiments, participants performed a dual task in which they had to study words presented visually while to-be ignored sounds were played over earphones. Sounds could be either standard or novel, and words could be presented in standard or novel font. Sounds were presented either simultaneously with the words (Experiment 1) or preceding them (Experiment 2). Electrophysiological correlates of novelty processing, the N2b and P3a ERP components, were recorded while the words were studied. It was seen that cued recall was better for words presented in novel fonts than for words in a standard font (the von Restorff effect). Words presented while novel sounds were played were remembered worse (Experiment 1) or equally well (Experiment 2) than those combined with standard sounds. Words presented in novel fonts elicited enhanced N2b, P3a, P3b, and N400 components; however, none of these components were specifically larger for subsequently recalled novel-font words. A larger N2b was found for recalled than for nonrecalled words, but this effect was not specific for words presented in novel font. We hypothesized that if novelty was beneficial for memory processing, the N2–P3 complex would be more enhanced for novel words that were later recalled than for those not recalled. The data showed otherwise. This suggests that novelty processing, as indexed by the N2–P3 novelty components, is not the main cause of the von Restorff effect.