On the basis of relevance theory’s claim that the relevance principle underlies all forms of communication, Chapter 3 examines how the RT model can be applied to communication involving visuals, and what adaptations are called for to achieve this goal. After reflecting on what constitutes visual communication, and showing how static visuals often combine with written language to create multimodal meaning, all of the RT concepts discussed in Chapter 2 are reconsidered with reference to their pertinence to analyzing visuals. This reconsideration will not only benefit visual and multimodal theories but will also provide new angles on classic RT. Whereas many RT concepts function without any problem when applied to visuals, there are others that cannot straightforwardly be “translated” to the visual realm and therefore need adaptation. The problematic issues mainly result from the fact that visuals typically have a structure and depicted entities, but no grammar and vocabulary. This leads to the questions of whether visuals can nonetheless be “coded”—which in turn has consequences for their possible underlying “logical form”—and whether information in visuals is necessarily always to be inferred or is sometimes actually decoded. Several examples are discussed to clarify these issues. In the final sections, there is a brief discussions of the relation between RT and Blending Theory, and of RT’s problematic take on metaphor.