This paper will demonstrate how we used state-level longitudinal data to model reading growth trajectories. Using data from large scale assessments that were vertically linked across grades in Hawaii, we utilized a multilevel regression framework to develop growth models to study students' reading performance trajectories during their elementary grades. The results of this study will provide a reference on the extent to which students' reading trajectories vary during their elementary school years and how this variation is related to student demographic characteristics. We also investigated and report how initial reading status is related to reading trajectories. The findings could provide practitioners with informed guidance on programmatic and student-level decisions based on reading performance from a longitudinal perspective.