The purpose of this research was to further investigate the information processing model suggested by Benjamin, McKeachie, Lin, and Holinger (1981) in explaining the poor academic performance of highly test-anxious students. Whereas previous studies pointing to deficits in the organization stage of processing mainly relied on self-reports, the first study of this research used a technique that enabled us to measure objectively the organization of course material by students in a nonevaluative situation. Results indicated that highly test-anxious students showed a disadvantage in organization of course material, in comparison with other students. In the second study, we investigated whether types of highly test-anxious students can be differentiated by their information processing skills. The first type includes those with good study habits who do not have problems in encoding and organizing the information but, rather, have a major problem in retrieval for a test. The second type includes those with poor study habits who have problems in all stages of processing. Such a distinction was supported by results comparing performance of these types of highly test-anxious students in an evaluative and a nonevaluative situation.Studies conducted in the last 30 years have generally supported the notion that high test anxiety is correlated with performance decrements (Benjamin, McKeachie, Lin, & Holinger, 1981;McKeachie, 1984;McKeachie, Pollie, & Spiesman, 1955). Numerous researchers have investigated the causes of these decrements in performance. According to traditional test-anxiety theory, a high anxiety level produces task-irrelevant responses, for example, error tendencies and worry that interfere with task-relevant responses necessary for good performance (Wine, 1971). As a result, much of the work on treatment strategies has been focused on the feelings and physiological reactions involved. Treatments aimed at relaxation or reducing emotional reactions to test situations, however, have had little effect on performance (Tryon, 1980). Desiderato and Koskinen (1969), Wittmaier (1972), and Culler and Holahan (1980) concluded that the poor academic performance of highly test-anxious students was due, at least in part, to inadequate study skills, resulting in poor knowledge of the relevant material. Benjamin et al. (1981), Tobias (1977), and Mueller (1980 suggested the use of an information processing model, which combines the aforementioned interference and study-habits models for the analysis of test anxiety. Such a model explains the performance deficit of highly test-anxious students in terms of problems in encoding and organizing information as well as in retrieval of this information in a test situation. To support their position, Benjamin et al. (1981) compared students with high, medium, and low test anxiety in an upper level psychology course.