Many gerontologists propose definitions of wisdom. Usually these are "empirical," as opposed to a-priori or "real" definitions. In this article we defend an a-priori definition of wisdom. We briefly explain a-priori and empirical definitions, and how they relate to each other in research. After rejecting two classical a-priori definitions of wisdom, we present and defend our own, and examine its ability to predict key findings of recent empirical studies. Finally, we describe some implications of our approach for future empirical studies of wisdom.