Cyberloafing is one of the phenomena that adversely affects the efficiency and productivity in learning and teaching activities in educational settings. Increased ICT (information and "communication technology") access status in educational environments and personal mobile devices lead to a wide range of cyberloafing behaviors of learners. In this regard, the aim of this study was to investigate cyberloafing behaviors of high school students in terms of several variables, including gender, ICT usage, unauthorized access to school network, metacognitive awareness and smartphone addiction. In this study, the relational screening model from descriptive research methods was used. A total of 269 9th grade students, 123 of whom were male and 146 of whom were female, were recruited. Hierarchical linear multilevel regression analysis was employed in the data analysis. The findings indicate that four of the five models designed to investigate cyberloafing behaviors in educational settings were found to be statistically significant. In other words, four hypotheses were supported. It has been concluded that the unauthorized access to the school network of learners has a significant impact on the cyberloafing behaviors in educational contexts. Additionally, the smartphone addiction and metacognitive awareness levels of students, as well as daily social media usage time are fundamental predictors of the cyberloafing behaviors.