Fictionalism in philosophy of science has had a second life when it was restated in terms of models. In this paper I first criticize Vaihinger’s conception of fictions as false but expedient assumptions. I argue that most of his examples of scientific fictions are either idealizations or conventions. I then criticize Fine’s brand of fictionalism, which is rooted on the pervasive practice of modelling and simulation in contemporary sciences. Fine conceives of models as useful fictions. I contend that whereas all scientific models are idealized to some degree, not all of them include fictional elements. I conclude that it is possible to regard scientific models as global representations of phenomena, even when they make use of fictions.