In most discourses of Art History, the art object is only present through representations. Art historians do see a lot of works of art, but when speaking, writing and reading about art and art history, that is, the standard routines for scholars and students, the works of art are absent more often that not. Does it make a difference to scholarship whether the subject of analysis is the original work of art or its reproduction? Three cases discuss this difference from a historiographical point of view. The first case is on Winckelmann and his use of reproductions, the next case on the color problem of Art History, and the third case on two art historians partly different reflections on the use of reproductions during the founding of Art History as a modern academic discipline.