The cognitive processes underlying children's multimedia text learning were studied using a self-paced task and authentic school texts with mostly representational pictures in an elementary school setting. Both the quantity and quality of learning were assessed immediately following intervention and 1 week later. In a within-subjects design, all of the children were taught four lessons of a different format: (1) written presentation only; (2) written presentation accompanied by pictures;(3) oral presentation only and (4) oral presentation accompanied by pictures. With respect to the quantity of learning, a short-term modality effect was found with oral presentation producing better results than written presentation when accompanied by pictures. A multimedia effect was found for only the oral conditions. With respect to the quality of the children's learning, an initial short-term modality effect was found but reversed 1 week following intervention with no evidence of a multimedia effect.