Although gemcitabine remains current first-line chemotherapy for patients with advanced pancreatic cancer, median survival times have not improved significantly since its introduction 15 years ago. Of the phase III trials which have investigated alternative regimens to single-agent gemcitabine, most have used combination regimens as the investigational arm. Accurate data on median overall, progression-free survival and objective response rates is important, for two principle reasons: advising patients about their prognosis and when powering phase II trials and evaluating the results of single-armed trials. This study aims to pool results from published randomized trials to date. Twenty-one randomized phase III trials involving a total of 6348 patients were identified from 1997 to 2010. Only one trial investigating a novel agent in combination with gemcitabine showed a significantly prolonged median and progression-free survival compared with single-agent gemcitabine. Pooled median and progression-free survivals for the single-agent gemcitabine arm involving 3171 patients across all studies were 6.15 and 3.3 months, respectively. Length of survival for patients with advanced pancreatic cancer remains disappointing. Further trials of novel agents to complement or replace gemcitabine are indicated.