The objective of this study is to construct a model which explains and predicts the relations of university students' cyberloafing behaviors with demographic and academic variables at computing courses where online social networking sites are utilized for education and is to review whether there is longitudinal effect on these relations in terms of learning experience. This group of the study is composed of 171 university students. In this study, self-description form, two different success tests, and various scales are utilized as data collection tools. For the analysis of the data, structural equation modeling and multiple hierarchical regression analysis are utilized. The results of this study indicate that variables of information technologies usage status, online learning activities usage status, academic self-efficacy, motivation, and cognitive absorption are predictable of cyberloafing behaviors at Time 2 point, and cyberloafing behavior predicts academic success and academic procrastination at Time 2 point. Although the relations between various study variables and cyberloafing are not meaningful at Time 1 point, it is an interesting finding that these relations are statistically meaningful at Time 2 point.