This paper reviews the differences between two interpretations accounting for the poor test performance of high anxious students: (1) that anxiety interferes with retrieval of prior learning; or (2) that, due to study skills deficits, the initial acquisition by anxious students is less thorough than by those lower in anxiety. Research results dealing with both hypotheses are reviewed. It is concluded that these are complementary, rather than mutually exclusive, formulations. A hypothesis is advanced predicting that test anxiety debilitates performance by reducing the cognitive capacity available for task solution, and that study skills facilitate learning by reducing the cognitive capacity demanded by different tasks. (Author)