In this report, the nature of anxiety is considered in historical perspective, and the concepts of trait and state anxiety are discussed. Methods for measuring these constructs are also described. In addition, hypotheses about the effects of anxiety on learning, formulated in terms of Spence-Taylor Drive Theory and Spielberger's Trait-State Anxiety Theory, are tested in two experiments on the effects of anxiety on computer-assisted learning. It was found in these studies that state anxiety was a better predictor of performance than trait anxiety, and that performance was an interactive function of A-State (State Anxiety) and task difficulty. Some important implications for the classroom teacher of research on anxiety and learning were discussed. (Author)