Integrating CYP2D6 genotyping and therapeutic drug monitoring (TDM) is crucial for guiding individualized atomoxetine therapy in children with attention-deficit/hyperactivity disorder (ADHD). The aim of this retrospective study was (1) to investigate the link between the efficacy and tolerability of atomoxetine in children with ADHD and plasma atomoxetine concentrations based on their CYP2D6 genotypes; (2) to offer TDM reference range recommendations for atomoxetine based on the CYP2D6 genotypes of children receiving different dosage regimens. This retrospective study covered children and adolescents with ADHD between the ages of 6 and <18, who visited the psychological and behavioral clinic of Children’s Hospital of Nanjing Medical University from June 1, 2021, to January 31, 2023. The demographic information and laboratory examination data, including CYP2D6 genotype tests and routine TDM of atomoxetine were obtained from the hospital information system. We used univariate analysis, Mann-Whitney U nonparametric test, Kruskal-Wallis test, and the receiver operating characteristic (ROC) curve to investigate outcomes of interest. 515 plasma atomoxetine concentrations of 385 children (325 boys and 60 girls) with ADHD between 6 and 16 years of age were included for statistical analysis in this study. Based on genotyping results, >60% of enrolled children belonged to the CYP2D6 extensive metabolizer (EM), while <40% fell into the intermediate metabolizer (IM). CYP2D6 IMs exhibited higher dose-corrected plasma atomoxetine concentrations by 1.4-2.2 folds than those CYP2D6 EMs. Moreover, CYP2D6 IMs exhibited a higher response rate compare to EMs (93.55% vs 85.71%, P = 0.0132), with higher peak plasma atomoxetine concentrations by 1.67 times than those of EMs. Further ROC analysis revealed that individuals under once daily in the morning (q.m.) dosing regimen exhibited a more effective response to atomoxetine when their levels were ≥ 268 ng/mL (AUC = 0.710, P < 0.001). In addition, CYP2D6 IMs receiving q.m. dosing of atomoxetine were more likely to experience adverse reactions in the central nervous system and gastrointestinal system when plasma atomoxetine concentrations reach 465 and 509 ng/mL, respectively. The findings in this study provided promising treatment strategy for Chinese children with ADHD based on their CYP2D6 genotypes and plasma atomoxetine concentration monitoring. A peak plasma atomoxetine concentration higher than 268 ng/mL might be requisite for q.m. dosing. Assuredly, to validate and reinforce these initial findings, it is necessary to collect further data in controlled studies with a larger sample size.