There is empirical evidence that self-explicated preference measurement methods are surprisingly robust in comparison to conjoint analysis. However, there has been no broad comparison of self-explicated methods with Choice-Based Conjoint analysis. The latter method gains more and more importance in marketing research practice. This empirical study shows that choice-based conjoint analysis leads to decisively better predictive validity than self-explicated measurement. Furthermore, its predictive validity is better than that of a new non-compensatory preference measurement method called RSS and the widespread Adaptive Conjoint Analysis (ACA).