Both gemcitabine (GEM) and fluoropyrimidines are valuable treatment for advanced pancreatic cancer. This open-label study was designed to compare the overall survival (OS) of patients randomly assigned to GEM alone or GEM plus capecitabine (GEM-CAP). Patients with previously untreated histologically or cytologically proven locally advanced or metastatic carcinoma of the pancreas with a performance status <or= 2 were recruited. Patients were randomly assigned to GEM or GEM-CAP. The primary outcome measure was survival. Meta-analysis of published studies was also conducted. Between May 2002 and January 2005, 533 patients were randomly assigned to GEM (n = 266) and GEM-CAP (n = 267) arms. GEM-CAP significantly improved objective response rate (19.1% v 12.4%; P = .034) and progression-free survival (hazard ratio HR, 0.78; 95% CI, 0.66 to 0.93; P = .004) and was associated with a trend toward improved OS (HR, 0.86; 95% CI, 0.72 to 1.02; P = .08) compared with GEM alone. This trend for OS benefit for GEM-CAP was consistent across different prognostic subgroups according to baseline stratification factors (stage and performance status) and remained after adjusting for these stratification factors (P = .077). Moreover, the meta-analysis of two additional studies involving 935 patients showed a significant survival benefit in favor of GEM-CAP (HR, 0.86; 95% CI, 0.75 to 0.98; P = .02) with no intertrial heterogeneity. On the basis of our trial and the meta-analysis, GEM-CAP should be considered as one of the standard first-line options in locally advanced and metastatic pancreatic cancer.