We studied the pharmacokinetics of 6-thioguanine (6TG) in 50 children treated for newly diagnosed acute myeloid leukemia, four of them with Down syndrome (DS). They received oral 6TG 100 mg/m2 body surface area twice daily for 4 days. Etoposide, 100 mg/m2/24 h, and cytarabine, 200 mg/m2/24 h, were administered concomitantly by intravenous infusion. On day 5, doxorubicin 75 mg/m2 was given as an 8-h infusion. The concentration of thioguanine nucleotides (TGN) in erythrocytes, the active metabolites of 6TG, was determined by high-performance liquid chromatography. The mean TGN concentration from 72, 95, and 106-h samples was used as a measure of drug exposure for each individual. The median TGN concentration in non-DS children above 2 years of age was 2.30 μmol/mmol Hb (range 0.57–25.3). The TGN concentrations varied widely (30-fold) also after dose normalization. We found no correlation with demographic, clinical, or biochemical parameters, and differences in bioavailability might be the most important explanation to interpatient variability. Children with high TGN concentration tended to have longer treatment interval to the next course, but we found no correlation with our predefined parameters for clinical response, that is, remission and relapse rate. Therefore, 6TG does not seem to be a candidate for therapeutic drug monitoring by TGN measurement, at least not in the setting of short multidrug treatment courses. Children with DS had significantly higher TGN concentrations, indicating that dose reduction might be considered to reach the same drug exposure as in non-DS children.