AIM: To evaluate the dose-limiting toxicities(DLTs)and determine the maximum-tolerated dose(MTD) and recommended dose(RD) of combination chemotherapy with gemcitabine, cisplatin and S-1 which is an oral fluoropyrimidine pro-drug in patients with advanced biliary tract cancer.METHODS: Patients with histologically or cytologically confirmed unresectable or recurrent biliary tract cancer were enrolled. The planned dose levels of gemcitabine(mg/m2), cisplatin(mg/m2), and S-1(mg/m2 per day) were as follows: level-1, 800/20/60;level 0, 800/25/60; level 1, 1000/25/60; and level 2,1000/25/80. In each cycle, gemcitabine and cisplatin were administered intravenously on days 1 and 15,and S-1 was administered orally twice daily on days 1to 7 and days 15 to 21, every 4 wk.RESULTS: Twelve patients were enrolled, and level0 was chosen as the starting dose. None of the first three patients had DLTs at level 0, and the dose was escalated to level 1. One of six patients had DLTs(grade 4 febrile neutropenia, leucopenia, and neutropenia; grade 3 thrombocytopenia) at level 1.We then proceeded to level 2. None of three patients had DLTs during the first cycle. Although the MTD was not determined, level 2 was designated at the RD for a subsequent phase Ⅱ study.CONCLUSION: The RD was defined as gemcitabine1000 mg/m2(days 1, 15), cisplatin 25 mg/m2(days1, 15), and S-1 80 mg/m2 per day(days 1-7, 15-21),every 4 weeks. A phase Ⅱ study is planned to evaluate the effectiveness of combination chemotherapy withgemcitabine, cisplatin, and S-1 in advanced biliary tract cancer. AIM: To evaluate the dose-limiting toxicities (DLTs) and determine the maximum-tolerated dose (MTD) and recommended dose (RD) of combination chemotherapy with gemcitabine, cisplatin and S-1 which is an oral fluoropyrimidine pro-drug in patients with Advanced biliary tract cancer. METHODS: Patients with histologically or cytologically confirmed unresectable or recurrent biliary tract cancer were enrolled. The planned dose levels of gemcitabine (mg / m2), cisplatin (mg / m2), and S-1 day) were as follows: level-1, 800/20/60; level 0, 800/25/60; level 1, 1000/25/60; and level 2, 1000/25/80. In each cycle, gemcitabine and cisplatin were administered intravenously on days 1 and 15, and S-1 was o orated twice daily on days 1to 7 and days 15 to 21, every 4 wk .RESULTS: Twelve patients were enrolled, and level 0 was chosen as the starting dose. None of the first three patients had DLTs at level 0, and the dose was escalated to level 1. One of six patients had DLTs (grade 4 febrile neutrope nia, leucopenia, and neutropenia; grade 3 thrombocytopenia) at level 1. We then proceeded to level 2. None of three patients had DLTs during the first cycle. Although the MTD was not determined, level 2 was designated at the RD for a subsequent phase II study. CONCLUSION: The RD was defined as gemcitabine 1000 mg / m2 (days 1, 15), cisplatin 25 mg / m2 days1 and 15 and S-1 80 mg / m2 per day days 1-7 -21), every 4 weeks. A phase II study is planned to evaluate the effectiveness of combination chemotherapy with gemcitabine, cisplatin, and S-1 in advanced biliary tract cancer.