Learning is increasingly representationally mediated; we rarely learn from direct experience but from representations that can be considered to encapsulate certain aspects of a phenomena. Textbooks and ebooks, simulations and games, augmented reality and haptic trainers provide innovative, often interactive forms, of representations in schools and universities, homes, professional training and museums. Moreover, professional activities are understood to be representationally focused. Experts select and construct representations appropriate for the situation, explanation or argument and argue from and for the representations in ways that make sense to others in their community. This chapter reviews a selective history of research on multi-representational learning, argues that this research this has resulted in a mature field of inquiry which has answers to some of the important questions involved in such learning and suggest future directions for research.