In this study, the authors explored a selection of middle-grade multimodal novels in order to push back against prior scholarship that tended to dismiss images in novels as mere decorations, visual reiterations of informational ready provided through written language (Godfrey, 2012). Inspired by the notion that the format of the middle-grade novel is becoming increasingly multimodal in nature (Serafini & Blasingame, 2012), the authors asked the following questions: (1) How have selected novels for middle-grade children deployed multiple modes of representation and communication in service of the story being told? (2) How have multimodal novels challenged or altered our conceptions about what constitutes a novel? (3) How do different modes work together within these texts to offer meaning potential? (4) How has the traditional conception of the novel as a language-dominant narrative been challenged by these other modes of representation? To address these questions, the authors systematically analyzed a data corpus of 34 middle-grade multimodal novels using an analytical tool they designed. A variety of perspectives supported by the theoretical framework guided the construction of the analytical tool. In the findings section, the authors assert that the relationship between visual images and written language within the data corpus novels can be described in one of three ways: as visual enhancement of written language, as visual and verbal construction of narrative, or as word-image fragmentary construction of narrative.