With greater prevalence of Internet access, there is an alarming trend in the number of students using the Internet in the classroom for non-class-related purposes. Cyberloafing (defined as personal Internet use at work or during class) has been extensively studied by researchers in workplace settings but not in education settings. Particularly, there is lack of research on developing a valid and reliable scale to measure cyberloafing behaviour among students. Hence, this study aims to examine the prevalence of cyberloafing activities among university students and to validate the cyberloafing scale of Akbulut et al. (2016) in the Malaysian context. A total of 238 usable data was collected from the 30-item cyberloafing scale that assessed five dimensions of cyberloafing behaviour namely sharing, shopping, real-time updating, access to online content and gambling / gaming. Descriptive analysis shows that students spend more time on sharing- related activities and least time on gambling / gaming-related activities in the classroom. Based on exploratory factor analysis, five factors are retained with most of the items loaded on its intended dimension factors, suggesting evidence of construct validity. The analysis also indicates that convergent validity is achieved as the factor loadings of each set of items measuring its intended dimension factors are above 0.5. Given that the correlations between extracted factors are not highly correlated, discriminant validity is warranted. These results support the investigated cyberloafing scale as reliable and valid.