Neuroimage studies have reported functional connectome abnormalities in posttraumatic stress disorder (PTSD), especially in adults. However, these studies often treated the brain as a static network, and time-variance of connectome topology in pediatric posttraumatic stress disorder remain unclear. To explore case-control differences in dynamic connectome topology, resting-state functional magnetic resonance imaging data were acquired from 24 treatment-naïve non-comorbid pediatric posttraumatic stress disorder patients and 24 demographically matched trauma-exposed non-posttraumatic stress disorder controls. A graph-theoretic analysis was applied to construct time-varying modular structure of whole-brain networks by maximizing the multilayer modularity. Network switching rate at the global, subnetwork, and nodal levels were calculated and compared between posttraumatic stress disorder and trauma-exposed non-posttraumatic stress disorder groups, and their associations with posttraumatic stress disorder symptom severity and sex interactions were explored. At the global level, individuals with posttraumatic stress disorder exhibited significantly lower network switching rates compared to trauma-exposed non-posttraumatic stress disorder controls. This difference was mainly involved in default-mode and dorsal attention subnetworks, as well as in inferior temporal and parietal brain nodes. Posttraumatic stress disorder symptom severity was negatively correlated with switching rate in the global network and default mode network. No significant differences were observed in the interaction between diagnosis and sex/age. Pediatric posttraumatic stress disorder is associated with dynamic reconfiguration of brain networks, which may provide insights into the biological basis of this disorder.