Advances in information technology have driven the adoption of online learning models in higher education institutions; however, they have also increased the potential for cyberloafing among students. This study aims to analyze the effects of online learning models, academic stress, and self-management on cyberloafing among active students at the Faculty of Islamic Economics and Business, State Islamic University of North Sumatra (FEBI UINSU), both partially and simultaneously. This study employs a quantitative approach with an explanatory research design. The study population consists of 4,089 active students at FEBI UINSU from the 2022–2025 cohorts, with a sample of 98 respondents determined using the Slovin formula and simple random sampling. Data were collected by distributing a questionnaire using a five-point Likert scale. Data analysis was carried out with IBM SPSS Statistics, applying validity tests, reliability tests, descriptive statistical analysis, classical assumption tests, multiple linear regression analysis, t-tests, F-tests, and the coefficient of determination (R²) test. The findings show that the online learning model exerts a positive and significant influence on cyberloafing (Sig. = 0.036 < 0.05), while academic stress also has a positive and significant influence on cyberloafing (Sig. = 0.000 < 0.05). Conversely, self-management shows no significant influence on cyberloafing (Sig. = 0.215 > 0.05). Taken together, the online learning model, academic stress, and self-management significantly influence cyberloafing (Sig. = 0.000 < 0.05). The coefficient of determination (R²) value of 0.533 shows that the three independent variables explain 53.3% of the variation in cyberloafing behavior, whereas the other 46.7% is driven by factors beyond the research model. These results indicate that academic stress is the most dominant factor affecting students’ cyberloafing behavior.