OBJECTIVE To explore the DNA incorporation of 6-thioguanine nucleotide levels (DNA-6TGN) during 6-mercaptopurine (6MP) therapy of childhood acute lymphoblastic leukaemia (ALL) and non-Hodgkin lymphoma (NHL) and its relation to erythrocyte levels of their metabolites: 6-thioguanine-nucleotides (E-6TGN), methylated metabolites (E-MeMP), Methotrexate polyglutamates (E-MTX), and to thiopurine methyltransferase activity (TPMT). METHODS We studied these metabolites in 229 blood samples from 18 children with ALL (N = 16) or NHL (N = 2) on 6MP/Methotrexate maintenance therapy. RESULTS DNA-6TGN levels were significantly correlated to E-6TGN (r (p) = 0.66, p = 0.003) with a trend to reach a plateau at high E-6TGN levels. To explore the relative DNA incorporation of 6TGN in relation to cytosol 6TGN levels, a DNA-6TGN index was calculated as DNA-6TGN/E-6TGN. The DNA-6TGN index was inversely correlated to E-6TGN (r (p) = -0.58, p = 0.012), which implies that with increasing levels of E-6TGN relatively less 6TGN are incorporated into DNA. E-MeMP levels were correlated to the DNA-TGN index (r (p) = 0.60, p = 0.008), indicating that high levels of MeMP result in enhanced DNA-6TGN incorporation, possibly due to inhibition of purine de novo synthesis, mediated by some of the methylated 6MP metabolites. CONCLUSIONS DNA-6TGN may prove to be a more relevant pharmacokinetic parameter for monitoring 6MP treatment intensity than the previously used erythrocyte 6MP metabolites levels. Prospective clinical trials are needed to evaluate the usefulness of DNA-6TGN for individual dose adjustments.