BackgroundPesticides may impact children’s neurodevelopment. As children’s metabolic function and neural plasticity change throughout their growth and development, the effects of pesticide exposure may also vary. ObjectivesWe aimed to identify the trajectories of combined pesticide exposure during childhood, and examine the associations of the exposure trajectories with children’s neurobehavior at the age of 10. MethodsWe involved repeated measurements of three pesticide metabolites [Pentachlorophenol (PCP), 3,5,6-Trichloro-2-pyridinol (TCPy), and Carbofuran phenol (CFP)], in urine samples collected from children in a cohort study at ages 1, 2, 3, 6, 7, 8, 9, and 10 years. The group-based multi-trajectory model (GBMT) and latent class analysis (LCA) were separately utilized to describe the distinct trajectories and patterns of pesticide mixture exposure during childhood. Meanwhile, the Strengths and Difficulties Questionnaire (SDQ) and attention deficit hyperactivity disorder (ADHD) Criteria of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) list were applied to assess behavioral disorders in children. The associations between exposure trajectories and behavioral problem scores were then examined. ResultsThe GBMT model delineated three distinct trajectories of combined pesticide exposure among children: consistently low, higher levels in early childhood transitioning to lower levels during pre-school age, and lower levels in early childhood followed by higher levels in the middle childhood. The LCA model identified three similar longitudinal exposure patterns. Further, the children in the second trajectory group identified by GBMT, characterized by higher early childhood exposure levels, exhibited significantly elevated hyperactivity/inattention scores of the SDQ compared to the other two groups (β = 0.46, 95 %CI: 0.11, 0.81; β = 0.44, 95 %CI: 0.02, 0.86). ConclusionsOur study revealed that exposure to pesticides during early childhood (especially before the age of two), rather than other age periods, was linked to hyperactivity/inattention problems in children aged 10 years. We also provided a novel perspective on characterizing the fluctuation in repeated measurements of multiple environmental chemicals and identifying the potential critical windows.