Why is it that students of apparently equal capability vary greatly in their ability to learn to use a computer system, that some individuals require little more than a manual while others require careful explanation before they are able to complete even the simplest tasks? To answer this question an experiment was conducted where students were taught how to use a computerized beverage management system by one of two instructional delivery methods, manual-based training (MBT) and instructor-based training (IBT). Subjects trained using IBT took significantly longer to train than did subjects using MBT. After being trained, however, subjects using MBT took significantly longer to complete a series of problems than did subjects using IBT. Correlations between instructional method and performance time were partially explained by the individual characteristic of field-dependence/ independence. The characteristics of anxiety, prior computer experience, sex and age did not significantly relate to training delivery method or performance. Results like these that relate success of specific training delivery methods to characteristics of individual learners may suggest ways to change computer training to accommodate a wider range of users.