Background : Prophylactic voriconazole use is recommended in children undergoing hematopoietic cell transplantation (HCT). Dosing considerations are essential due to its narrow therapeutic index. Known covariates do not sufficiently explain large interindividual pharmacokinetic (PK) variability of voriconazole. Moreover, knowledge of voriconazole PK for age <2 years is limited. Objectives : We investigated genetic and clinical covariate association with voriconazole interindividual PK variability and subsequently simulated dosing regimens in children. Methods : This study was conducted as a part of a single-institution, phase I study of intravenous voriconazole in children undergoing HCT. We conducted a population PK analysis and tested covariate effects on voriconazole PK, including 67 genetic variants and clinical variables. Results : We analyzed plasma voriconazole and n-oxide metabolite concentrations from 58 children aged <21 years (n=12 in age <2 years). A two-compartment parent mixed linear/nonlinear model best described our data. CYP2C19 phenotype and body weight were significant covariates (both p<0.05). Our model performance in age <2 years was comparable to other age groups. Simulation of the final model suggested the following dosages to attain target steady-state trough concentrations of 1.5 - 5.0 mg/L in CYP2C19 normal phenotype: 16 mg/kg (weight <15 kg), 12 mg/kg (weight 15-30 kg), 10 mg/kg (weight >30 kg), whereas dosages were 33-50% lower for CYP2C19 poor/intermediate and 25-50% higher for CYP2C19 rapid/ultrarapid phenotypes. Conclusions : We propose a new starting dosage regimen, combined with therapeutic drug monitoring for intravenous voriconazole in children of all ages. Future studies should validate this dosing regimen.