Simulation is used to study manufacturing systems, mainly for the purpose of problem solving or decision making. To make sure that the simulation model is able to provide a result that can be trusted, the model needs to be validated. The purpose of this paper is to argue that not all simulations need a completely valid simulation model. Validation aims at making the model as valid as possible. Some validation methods are more suited for certain simulation applications, while others can be less appropriate. The aim is to always use a validation method that needs the least modeling effort, but that guarantee a valid simulation result. The contribution of this paper is a model that describes the connections between simulation applications and the degree of validity needed for each application. The most common simulation applications are represented in the model and they are given a position on the validity-scale. The position recommends the lowest degree of validity that the simulation model must have, in order to provide a valid simulation result. This will reduce the time and the cost of simulation studies.