Considering the demanding higher education context, university students are at risk to experience burnout symptoms such as emotional exhaustion, cynicism, and reduced professional efficacy. Theoretical models (e.g., Job Demands-Resources Model) state that burnout symptoms develop over time, vary between individuals, and bear adverse consequences. To date, the temporal development of student burnout symptoms within semesters, inter-individual differences in their development, and the role of their development for academic outcomes like student dropout are understudied. To complement this limited research, we used date of a three-wave longitudinal study with N = 1435 undergraduate students over one semester. First, we modeled unconditional second-order latent growth curves to examine the initial levels and trajectories of each burnout symptom (emotional exhaustion, cynicism, and reduced professional efficacy). We then modeled conditional second-order latent growth curves to examine whether initial levels and trajectories differed depending on students’ sociodemographic characteristics (gender, age, study progress, and academic major) and associated with students’ intentions to drop out from university. Results indicated a linear increase in student burnout symptoms over the semester. Sociodemographic differences existed in the initial levels of emotional exhaustion and reduced professional efficacy and the trajectories of each symptom. Generally, female, STEM, and higher- and lower- semester students were particularly affected by burnout symptoms. Further, higher initial levels and an increase in each symptom corresponded with higher dropout intentions. Summarizing, the results highlighted that student burnout symptoms increase as the semester progresses, seem more pronounced among certain student groups, and operate as strong predictors for dropout intentions.