Within the recognized tensions between statewide testing and the process of teaching and learning, formative assessment's potential for improving student learning and for shedding light "inside the black box," has received increased attention from scholars in different countries. In their critical review, Dunn & Mulvenon (2009) pointed out the lack of agreed-upon definitions and limited empirical evidence concerning formative assessment. We contend that from the educational assessment field, there is a clear overlap among definitions of the concept and that the conceptual confusion may be coming from outside the field. We also argue that Dunn & Mulvenon's claim of limited empirical evidence is basically inaccurate and based on misinterpretations of Black & Wiliam's (1998) article and of social science research more broadly. Consequently, we start by distinguishing key concepts related to formative assessment. Second, we describe the paradigm of the educational assessment field, and summarize the different definitions they propose. Finally, the article addresses the issue of "limited" empirical evidence concerning formative assessment. (Contains 1 table.)