The purpose of this research is to examine the relationships between changes to computer self-efficacy (CSE) and computer anxiety and the impact on performance on computer-related tasks in both online and face-to-face mediums. While many studies have looked at these factors individually, relatively few have included multiple measures of these factors concurrently and assessed the effects over a period of time in a realistic environment. Transactional Distance Theory and previous research are used to develop a theoretical model integrating these factors. Transactional distance can be viewed as an outcome of the culture within an educational environment. As such the three transactional distance variables--structure, dialogue, and learner autonomy--may be measured by certain aspects of that culture. Describing the two mediums in terms of transactional distance allowed us to explore the interrelationships between the changes in CSE, Anxiety, and Performance in face-to-face and online classes. Data are drawn from students in a junior level Management Information Systems (MIS) class at a medium-size public University in the Southeast U.S.A. Classes in both mediums completed a real life computer-based project in which previously validated instruments are used to longitudinally measure perceptions of computer self-efficacy (CSE), computer anxiety, and transactional distance. Partial Least Squares (PLS), a structural equation modeling technique, is used to com-pare the results of these factors on performance in both the face-to-face and online mediums. The results show transactional distance was indeed acting as an anxiety-reducing mechanism. A comparison of differences between mediums suggests that the various aspects of transactional distance influence each differently. The findings highlight the importance of structure and innovation in the online medium while aspects of dialog were shown to be important in the face-to-face medium. In effect, the interactive or emotive characteristics of transactional distance were more important in the face-to-face medium, while the structure and individual autonomy aspects were more influential in the online medium. Thus, one implication for face-to-face instructors is the need to continually maintain free-flowing dialog with the students to enhance learning. On the other hand, logical organization is very important in an online medium. Finally, the significance of the innovation component illustrates the importance of using new techniques and technologies to improve the interaction aspect of online education. Understanding these differences and implementing measures to accommodate these differences could increase teaching effectiveness and ultimately result in improvements in performance on computer-related tasks. (Contains 1 table and 4 figures.)