Symptoms of chronic conditions that patients experience are not only multidimensional but also longitudinal. Advanced analytic techniques are needed to examine the two types of symptom characteristics simultaneously to understand the dynamics of trajectories of multiple symptoms. Unfortunately, most symptom studies in the literature focus on either individual symptoms or clusters of symptoms that involve cross-sectional data. This study utilized an innovative modeling technique, multitrajectory modeling in latent class growth analysis, to analyze five interrelated symptoms (fatigue, sleep disturbances, pain, nausea, and depression) in 236 children with leukemia over a 12-month therapy period. Multitrajectory modeling can help identify linkages between the trajectories of multiple symptoms over time. Each trajectory class is defined not by one symptom but by multiple symptoms simultaneously. In this study, three latent classes of children following similar patterns of symptom trajectories were identified: constantly mild, constantly severe, and slightly decreased. The three classes can be modeled in subsequent analyses as a clinically meaningful predictor of patient outcomes.