Our minds are capable of perceiving similarities not only within the same but also across different sensory areas and different cognitive domains. Iconicity is representation based on similarity, and cross-modal iconicity, which is an extremely widespread phenomenon, should be understood as iconicity that crosses the borders of different kinds of material, spatiotemporal, and sensorial modes, and, furthermore, the border between sensory structures and cognitive configurations. For instance, a visual entity may resemble and thus iconically represent something that is auditory or abstractly cognitive. The aim of this semiotic study, substantially based on empirical findings in psychological, cognitive, and neurological research, is to suggest a general theoretical framework for conceptualizing cross-modal iconicity and relating different kinds of mono-modal and cross-modal iconicity to each other. Its chief argument is that perception and conception of images and metaphors should be understood as the two extremes in a continuum of iconic representation where cross-modal iconicity bridges the apparent gap between mono-modal, sensory-based iconicity and cognitive iconicity. It offers an outline of material, spatiotemporal, and sensorial modes and their interrelations; a thorough account of cross-modal iconicity; a conceptual structure for charting degrees of similarity and iconicity with the aid of cross-modality; an array of examples illustrating the continuum of iconicity from image to metaphor; and a brief discussion of the notion of image schema as an explanatory factor for cross-modal iconicity.