We previously presented an experimental model of Barrett's adenocarcinoma of the esophagus by demonstrating that esophagojejunostomy combined with subcutaneous injection of 2,6-dimethylnitrosomorpholine in Sprague-Dawley rats resulted in development of adenocarcinoma in the distal esophagus. The present study was devised to investigate the influence of pancreatic and biliary duodenal-content reflux on the induction of esophageal carcinoma. Three groups of 8-week-old Sprague-Dawley rats were controls: the first was exposed to pancreatic reflux, the second to biliary reflux, and the third to both. The other three experimental groups were similar except that a 1/100 LD50 dose of 2,6-dimethylnitrosomorpholine was injected subcutaneously weekly, starting on day 15. Carcinoma of the esophagus was induced only in animals receiving the carcinogen after exposure to either pancreatic reflux (3/22, 13%) or pancreatic and biliary reflux (9/27, 33%). Half of the carcinomas were adenocarcinoma and half were squamous cell carcinoma. These findings suggest that under these experimental conditions, in which the carcinogen is used in a low dose, esophageal carcinoma is induced only when pancreatic secretions are present in the duodenal-content reflux. Biliary reflux, however, appears to exert a cocarcinogenic effect when combined with pancreatic secretions. The clinical relevance of these findings needs further evaluation. Conceivably, the elimination of pancreatic and biliary duodenal-content reflux in patients with documented Barrett's mucosa may inhibit the progression from metaplasia to adenocarcinoma.