In recent years the history and philosophy of computer simulation has become an active research area. Practices of computer simulation have been described as “qualitatively diff erent ways of doing science” (Fox Keller 2003: 202), a “signifi cant and permanent addition to the methods of science” (Humphreys 2004: 64), a “revolution in science” (Baker et al. 1977; Schweber & Wächter 2000; Dahan & Aubin 2002), or a “third way in science” (Kaufmann & Smarr 1993; Galison 1996). Also, signifi cant problems of computer simulation have been described, particularly the problem of model validation and uncertainty (Oreskes et al. 1994; Heymann & Kragh 2010; Petersen 2006). The novelty of the methods and the epistemic problems involved in computer simulation also raise the question of how confi dence in models and simulations emerged. So far, surprisingly little research has been done on this question. In this chapter I will approach this question for the case of climate simulation in its early period from the mid1950s to about 1980. During this period climate modelling and simulation became an infl uential research fi eld. It rapidly and radically changed the meaning of and interest in climate and drew attention to climate change (Weart 2010; Heymann 2009, 2010a).