Research on the influence of attitudes toward computers and end user performance has reported inconsistent results. The inconsistent results, at least in part, could be attributed to the lack of correspondence between the general nature of the attitude measure and the specific nature of the criterion, end user performance. Based on Ajzen and Fishbein’s (1980) behavioral intentions model, we argue that attitudes toward working with computers matches end user performance in terms of specificity and relevance, and therefore should be consistently related to end user performance. In this study, in addition to attitudes toward working with computers, the effects of goal setting and self-efficacy on end user performance were also tested. Results indicate that attitudes toward working with computers, goal setting and self-efficacy significantly influence end user performance. Strong support for attitudes, goal setting and self-efficacy indicate that end user performance can be substantially enhanced by shaping end users’ attitudes toward working with computers, teaching end users to set specific and challenging goals, and enhancing end users’ beliefs to effectively learn and use computing technology.