The highly dynamic development of simulation technologies is propelled by the expectation that increasingly high-performing forecasting instruments can and will be employed. In current discussions, reference to "high-performing forecasting instruments' combines two perspectives that stand in an unresolved relationship to one another, which is philosophically revealing: forecasts as true, as in adequate, representations versus performance measured by the success of the technical practice. While the first perspective presupposes a theory of truth based on realism (adequate representations), the second orients itself towards pragmatic representations of truth. Once this is made explicit, a shortcoming in the existing philosophy of simulation becomes evident. An intense debate on the verification and validation of simulations has failed to address the theory of truth. This article undertakes a discussion on a theory of truth suitable for computer simulation that is not only based on a theoretical interest, but also on a practical one.