Relational reasoning, the ability to discern meaningful patterns within otherwise unconnected information, is regarded as central to human learning and cognition and as particularly critical for those functioning in today's information age. However, the literature on this foundational ability is currently housed within a range of domains of inquiry, where divergent terminology and methodologies are commonplace. This dispersion has made it difficult to harness the power of existing work to inform future research or guide educational practice. In order to address this lack of consolidation, a systematic review of relational reasoning was undertaken. Specifically, 109 empirical studies dealing with relational reasoning in general or one of four manifestations (i.e., analogy, anomaly, antinomy, and antithesis) were analyzed. Resulting data revealed trends across fields of inquiry, including a degree of conceptual ambiguity, conceptual and operational misalignment, and a lack of ecological validity in certain research paradigms. There were also particular forms and measures of relational reasoning that were more commonly investigated, as well as certain domains that were more often studied. Implications for how future research can examine relational reasoning as a multidimensional construct within educational contexts are also discussed.