The study investigates possible factors affecting Turkish students' perception of task difficulty in the Program for International Student Assessment (PISA) 2018. The research data consist of the survey responses of the students who participated in the PISA 2018. The participants include 6679 15-year-old Turkish students. The dependent variable of this study is the students' views of the reading task difficulty in the PISA 2018. Independent variables are divided into two as student-level and school-level variables. The student-level variables include strategy use, reading format preference, time spent reading, perceived difficulty, and frequency of being online. The school-related variable is schools’ digital capacity. Since the PISA sample has a nested structure, the data were analyzed using hierarchical linear modeling (HLM). The analysis showed that the differences among schools predicted approximately 9.8% of the variation in students' perception of the difficulty of reading literacy tasks in PISA while the rest of the percentage was due to the differences among students. In the regression model, student-level and school-level variables as groups appeared to predict the dependent variable significantly. In the final interactive regression model, the interactions of student and school-level variables were examined, and the interaction appeared not to be a significant predictor of the dependent variable. Based on the HLM analysis, it can be argued that PISA tasks are perceived as easier in cases where schools' digital capacity increases, students opt for strategy use, the use of electronic media in reading increases, the difficulty in reading and comprehension decreases, the time spent reading increases, and the frequency of being online increases.