Decision making, based on forward looking deliberation, assumes a cognitive representation of the causal relationships determining the consequences of behavioral choices. We refer to this perceived causal structure as the mental model which, of course, can — and in all complex environments — must differ from the actual decision situation. Mental modeling is thus one step in the possible recursive process of boundedly rational reasoning which precedes belief and goal formation and the search for satisficing behavioral options. How we cognitively perceive a choice task is usually not revealed by choice data. We discuss experimental methods trying to elicit more or less directly crucial aspects of mental modeling and how they might help to explain more reliably choice making. According to the few experimental studies challenging this inherently difficult task none of the experimental methods is without flaws. We nevertheless propagate using them.