Situational judgment tests (SJTs) are typically conceptualized as contextualized selection procedures that capture candidate responses to a set of relevant job situations as a basis for prediction. SJTs share their sample-based and contextualized approach with work samples and assessment center exercises, although they differ from these other simulations by presenting the situations in a low-fidelity (e.g., written) format. In addition, SJTs do not require candidates to respond through actual behavior because they capture candidates' situational judgment via a multiple-choice response format. Accordingly, SJTs have also been labeled low-fidelity simulations. This SJT paradigm has been very successful: In the last 2 decades, scientific interest in SJTs has grown, and they have made rapid inroads in practice as attractive, versatile, and valid selection procedures. Contrary to their popularity and the voluminous research on their criterion-related validity, however, there has been little attention to developing a theory of why SJTs work. Similarly, in SJT development, often little emphasis is placed on measuring clear and explicit constructs. Therefore, Landy (2007) referred to SJTs as "psychometric alchemy" (p. 418).To shed light on these pressing issues, this focal article builds a case for reconceptualizing SJTs as measures of a form of general domain knowledge.