Over the last 20 years there has been a proliferation of research on metacognition and its connection to specific variables. Questions remain regarding what combinations of subcomponents of metacognition contribute to the development of metacognitive expertise (ME) in high achievers, whether a multicomponential model of ME can be delineated and proven through additional research, and whether a scale can be developed to assess learners' ME so as to identify low achievers (at-risk learners). To answer these questions the researcher (a) designed a new model of ME by exploring Metacognitive Domains and Component Factors; (b) developed a scale, the Metacognitive Expertise Assessment Tool (ME-AT), and tested its reliability in comparison to the well-respected MAI; and (c) established the ME-AT's value as a predictive measure of ME. Overall, the findings revealed that three of the Metacognitive Domains and all 10 of the Component Factor predictors significantly discriminated low achievers from high achievers, whereas the MAI did not significantly discriminate low achievers from high achievers.