BACKGROUND A liquid formulation of 6-mercaptopurine (6-MP) was recently approved by the Food and Drug Administration (Purixan®) based on bioavailability (BA) data from healthy adults. We examined the pharmacokinetics (PK) and BA of 6-MP in children with acute lymphoblastic leukemia (ALL) comparing a marketed tablet, two extemporaneously prepared liquid formulations, and data from the approved liquid formulation. METHODS Twenty-two children (6-17 years) participated in a randomized two-way, crossover study of two cohorts. Group 1 (n = 11; five males) received a 5 mg/ml liquid formulation and the marketed 50 mg 6-MP tablet on separate occasions, and Group 2 (n = 11; five males) received a 50 mg/ml liquid formulation and the marketed tablet. The usual prescribed 6-MP dose (25-115 mg/m2 ) was given after an 8-hr fast. Serial blood samples were collected over 8 hr postdose. Plasma 6-MP concentrations were determined using a good laboratory practice (GLP)-validated liquid chromatography-tandem mass spectrometry method. PK parameters were calculated using noncompartmental analysis and compared within and between cohorts, and thiopurine methyltransferase (TPMT) genotype was analyzed. RESULTS No patient had a TPMT genotype reflective of a poor metabolizer phenotype. Comparison of PK parameters between 5 and 50 mg/ml treatments revealed significant differences (P <0.05) in AUCN (where AUC is area under the curve), CmaxN , and Tmax . Comparisons within each group revealed significant differences in AUC0-∞ and Tmax in the 5 mg/ml group. CONCLUSIONS Pharmacokinetic profiles of 6-MP established in healthy adults with the approved liquid formulation may not reflect the PK profile in children with ALL. Formulation-specific differences in PK may significantly impact the dose-exposure profile in these children and must be considered.