Advance pancreatic cancer (unresectable or metastatic) is a traditionally poor prognostic disease with short survival time. In the past decades, gemcitabine was confirmed as the standard therapy for pancreatic cancer with modest anti-tumor efficacy, 6-11% of overall response rate and 5-6 months of overall survival time. Due to unsatisfactory efficacy of single agent with gemcitabine, some efforts were made to explore the efficacy and adverse effects of combination therapy. 5-FU was the chemotherapeutic agent that showed the benefit in combination with gemcitabine. In the six clinical trials of gemcitabine combined with 5-FU versus gemcitabine alone, the hazard ratio of overall survival(OS) in the combination therapy ranged from 0.79 to 1.05. The meta-analysis of the regimen of gemcitabine combined with 5-FU showed the improvement of overall survival rate by 10%(HR=0.9). The GEST study demonstrated the noninferiority of S-1 to gemcitabine and the significant survival benefit of gemcitabine plus S-1 with median OS of 10.1 months. For its tolerable toxicities and superior survival benefit, Gemcitabine plus S-1 has become the first-line therapy in patients with good performance status in Taiwan. On the other hand, nanoliposomal irinotecan with fluorouracil and folinic acid were tested in patients with gemcitabine-refractory metastatic pancreatic cancer, which successfully extended survival with a manageable safety profile. This regimen represents a new treatment option for the second-line therapy. The feasibility and efficacy of triplet chemotherapy for advance pancreatic cancer was explored. The regimens of gemcitabine-based triplet chemotherapy(GOFL, GOLF, FOLFUGEMOX, SLOG) were reported by different study groups. The National Health Research Institutes of Taiwan developed the SLOG regimen, consisting of biweekly gemcitabine(fixed rate infusion of 800 mg/m2), followed by 85 mg/m2 oxaliplatin on D1 and oral S-1 35 mg/m2 and 20 mg/m2 leucovorin, twice daily D1-D7, every two weeks. Of the 54 patients in phase I/II study, the overall response rate and disease-control rate was 37% and 67%, respectively. The median progression-free survival and overall survival was 7.6 months and 10.7 months, respectively. The efficacy of SLOG regimen is promising and its safety profile was favorable. Another triplet chemotherapy without gemcitabine was reported in 2010 ASCO annual meeting. The ACCORD 11 trial, FOLFIRINOX(irinotecan, oxaliplatin, 5-FU/LV) regimen for metastatic pancreatic adenocarcinoma, showed superior efficacy to gemcitabine alone. However, the grade 3/4 adverse events of FOLFIRINOX seemed to be the major concern(neutropenia 45.7%, fatigue 23.3%, vomiting 14.5%, diarrhea 12.7%) for clinical practice. In conclusion, combination chemotherapy shows favorable efficacy and tolerable adverse effects in patients with advance pancreatic cancer. FOLFIRINOX seems to have superior efficacy but severe myelosuppression limits its clinical use. Gemcitabine-based doublet and triplet chemotherapy seems to be the mainstay systemic therapy for advance pancreatic cancer in clinical practice.