Background: Therapeutic drug monitoring for busulfan is important to prevent adverse events and improve outcomes in stem cell transplantation. We investigated intravenous busulfan pharmacokinetics and evaluated the utility of limited sampling strategy (LSS) as a simple method to estimate the area under the concentration-time curve (AUC). Procedure: The study comprised 87 busulfan measurements in 54 children who received intravenous busulfan between August 2015 and May 2020. AUCs were calculated from 3–5 blood sampling points in each patient, and the correlation between AUC and plasma concentrations (ng/mL) at 1, 2, 3, 4, and 6 h after initiating busulfan infusion (C₁, C₂, C₃, C₄, and C₆, respectively). Results: By one-point sampling strategy, the most accurate predicted AUC was based on C₆ (r² = 0.789; precision, 11.0%) in all patients. The predicted AUC based on C₆ was highly precise (r² = 0.937; precision, 5.9%) in adolescent patients weighing > 23 kg, but the correlation was poor in infants and young children weighing ≤23 kg (r² = 0.782; precision, 11.4%). By two-point sampling strategy, the predicted AUC based on C₃ and C₆ showed the most favorable performance (r² = 0.943; precision, 6.4%), even in infants and young children, whereas the predicted AUC based on C₃ and C₆ was acceptable (r² = 0.963; precision, 5.7%). Conclusions: The AUC of busulfan can be predicted based on C₆ in adolescent patients. However, there was substantial inter-individual variation in busulfan pharmacokinetics in infants and young children, in whom two-point LSS was necessary for accurate AUC prediction.