This article argues that conducting experiments involving the ability to control and even manipulate the cognitive load in working memory (WM; storage and or processing load) should make it possible to identify the processes involved during the integration of information coming from multiple sources. Some experiments using the dual-task paradigm are reviewed, and an original experiment using complex multimedia material is presented. Overall, the experiments show that even in cases where subjects have to navigate between different types of information and have to integrate various items of information, the verbal storage component of WM is important in permitting comprehension. Visuospatial WM is in addition involved as soon as visuospatial processing is required. Storage of verbal information does not however depend on the modality of presentation and the classical modality effect appears to depend on individual differences. Some theoretical and practical implications of these results are drawn.In the context of the development of new technologies for the communication of information, multimedia systems are developing quickly and raise many questions, especially with reference to education. A multimedia system can be defined as a system which requires the integration of different types of information: verbal information presented visually or auditorily (e.g. words, sentences or short texts), pictorial information presented visually in a static or dynamic way (illustrations, photographs, diagrams) and sound information. In addition, systems that allow users to navigate between different sources of information through the use of hypertext structures are often considered to be multimedia systems, even if only one type of information is provided (for example, verbal information presented visually). To what extent does the use of such systems make a positive contribution to the learning process? There is a strong temptation to simply assume that using various information presentation formats, using realistic and vivid presentations and providing multiple possibilities for interaction with a learning system generally results in better learning. The acquisition of information via any technical system remains, however, subject to the constraints of human information processing. Existing models of multimedia comprehension originate from text comprehension models (e.g. Schnotz, 2005) or from research into the effect of illustrations (e.g. Mayer, 2001). One core concept APPLIED COGNITIVE PSYCHOLOGY