This article discusses the assessment of self-regulated learning processes as students acquire cognitive skills in specific academic domains. Domain-specific assessment is useful for understanding student learning and for planning instructional activities that help to promote it. Although much psychological research has used general measures of cognitive functioning to study human behavior, general measures do not reliably predict what people do in specific situations. Some measures of student aptitude, such as intelligence or abilities, correlate with student achievement, but complex aptitude constellations often predict learning better than any aptitude alone. The assessment of one type of domain-specific self-regulated learning process--perceived self-efficacy--is discussed to include instrument development, reliability, and validity. Empirical evidence is presented demonstrating the predictive utility of self-efficacy in various domains, along with research showing how self-efficacy is affected by instructional contexts. Future research suggestions are provided. (A six-page list of bibliographic references is attached.) (Author/JAZ)