OBJECTIVES: Metastatic esophageal carcinoma is an incurable disease with median survival duration of 6 to 8 months. Based on preclinical data suggesting a dose-dependent synergy between gemcitabine and irinotecan we have conducted a phase II trial in patients with advanced or metastatic esophageal carcinoma. METHODS: Patient eligibility included a diagnosis of squamous cell or adenocarcinoma of the esophagus/gastroesophageal (GE) junction, metastatic or recurrent disease, no CNS metastasis, no prior chemotherapy, prior adjuvant/neoadjuvant chemotherapy was allowed, no prior gemcitabine or irinotecan, performance status of 0 to 2 and adequate organ function. Patients received gemcitabine 1000 mg/m2 and irinotecan 100 mg/m2 given day 1 and day 8, every 3 weeks. The primary end point was the 6-month survival rate. The secondary end point was to assess qualitative and quantitative toxicities. RESULTS: Fifty-seven eligible patients were accrued. There were 4 treatment-related deaths. The primary grade 3 to 4 toxic events were diarrhea, dehydration, neutropenia, thrombocytopenia, anemia, and anorexia; and 4 episodes of grade 3 to 5 febrile neutropenia, 1 fatal. The study was designed to detect a difference between the null hypothesis of 30% 6-month survival and the alternative hypothesis of 50% 6-month survival. The Kaplan-Meier estimate of 6-month survival is 56% (95% CI: 43-69%), with a median of 6.3 months. The median time to progression was 3.7 months. The 6-month progression-free survival estimate is 25% (95% CI: 13-36%). CONCLUSIONS: The length of survival suggests that this combination has benefit similar to platinum containing regimens, however, the toxicity is substantial and is unlikely to prove superior to platinum containing regimens.