The Children's Cancer Group 1952 (CCG-1952) clinical trial studied the substitution of oral 6-thioguanine (TG) for 6-mercaptopurine (MP) and triple intrathecal therapy (ITT) for intrathecal methotrexate (IT-MTX) in the treatment of standard-risk acute lymphoblastic leukemia. After remission induction, 2027 patients were randomized to receive MP (n = 1010) or TG (n = 1017) and IT-MTX (n = 1018) or ITT (n = 1009). The results of the thiopurine comparison are as follows. The estimated 7-year event-free survival (EFS) for subjects randomized to TG was 84.1% (+/- 1.8%) and to MP was 79.0% (+/- 2.1%; P = .004 log rank), although overall survival was 91.9% (+/- 1.4%) and 91.2% (+/- 1.5%), respectively (P = .6 log rank). The TG starting dose was reduced from 60 to 50 mg/m(2) per day after recognition of hepatic veno-occlusive disease (VOD). A total of 257 patients on TG (25%) developed VOD or disproportionate thrombocytopenia and switched to MP. Once portal hypertension occurred, all subjects on TG were changed to MP. The benefit of randomization to TG over MP, as measured by EFS, was evident primarily in boys who began TG at 60 mg/m(2) (relative hazard rate [RHR] 0.65, P = .002). The toxicities of TG preclude its protracted use as given in this study. This study is registered at http://clinicaltrials.gov as NCT00002744.