["Journal of Computer Assisted Learning, Volume 42, Issue 3, June 2026. ", "\nABSTRACT\n\nBackground\nEngagement is an essential ingredient in achieving learning success. It can be understood as a dynamic process that unfolds over time, and is shaped by several factors related to the learner and the learning context. Understanding how learners fluctuate between engagement and disengagement is vital for many stakeholders. Especially in the case of first‐year bachelor's students who are prone to high attrition and low success rates, it is important to consider engagement or lack thereof longitudinally.\n\n\nObjectives\nThis study aims to analyse the longitudinal engagement trajectories of first‐year business students to inform interventions that can foster sustained engagement.\n\n\nMethods\nUsing the VaSSTra methodology, we explore the longitudinal engagement of students in the first year of a Business bachelor's program. A large dataset of 13 semester‐length courses of 651 students and 8463 student enrollments is examined.\n\n\nResults and Conclusions\nThree engagement states (active, average and disengaged) and three distinct longitudinal engagement trajectories (mostly engaged, mostly average and mostly disengaged) are identified. Stability and heterogeneity were explored using within‐student and between‐student entropy. The Mostly Engaged trajectory was stable with infrequent changes in engagement across courses and scored the highest. The Mostly Average trajectory was the most frequent with less stability and more transitions between states. The Mostly Disengaged trajectory became more disengaged in courses of the second semester. The results of our study show that interventions and support should be fortified in the first semester itself to support students.\n\n"]