X-in-the-Loop (XiL) tests are a key enabler to scale up testing width and depth as a supplement of Proving Ground (PG) tests for Advanced Driver Assistance Systems (ADAS). Since simulation models are major building blocks in XiL, the question of simulation validity is gaining importance. In this contribution, a validity argumentation for a concrete scenario executed in a XiL environment is proposed. Both tests in PG and XiL are prone to uncertainties. So-called aleatory uncertainties represent stochastic processes as they occur in PG and may also exist in XiL. In contrast, epistemic uncertainties reflect a lack of knowledge and are inherent in each XiL simulation model. We propose the following validity argumentation: A XiL environment is considered valid if its error falls within a tolerance band defined by the aleatory uncertainty of PG tests. Multiple validation metrics are computed for so-called System Response Quantities (SRQs). The method is applied to three concrete scenarios used to test an Advanced Emergency Brake System (AEBS), with XiL represented by a Software-in-the-Loop (SiL) environment.