To determine the clinical efficacy of Roux-en-Y reconstruction (RY) after distal gastrectomy, we compared postoperative outcomes of patients who underwent RY or conventional Billroth I reconstruction (B-I). A total of 50 patients were prospectively randomized to either B-I or RY reconstruction, and complications, postoperative course, and nutritional status were compared. Bile reflux and inflammation in the remnant stomach and lower esophagus were evaluated by postoperative follow-up endoscopy at 6 months. Operative time and blood loss as well as postoperative nutrition did not show significant differences between the two groups. As anticipated, 5 of 24 patients with RY reconstruction developed gastrojejunal stasis in the early postoperative period, which led to a longer postoperative hospital stay as compared with the B-I group (mean +/- S.D; B-I; 19.0 +/- 6.2, RY; 31.8 +/- 21.7 days) (P < 0.05). Endoscopic examination revealed that the frequency of bile reflux (P < 0.01) and degree of inflammation in the remnant stomach (P < 0.05) were less in the RY group than in the B-I group. However, inflammatory findings in the lower esophagus were observed in 7 (27%) of B-I, and 8 (35%) of the RY group, suggesting that late phase esophagitis was not improved in the RY group. Roux-en-Y reconstruction was effective in preventing duodenogastric reflux and resulting gastritis, but it did not prevent esophagitis. Because RY reconstruction induces the frequent complication of Roux-en-Y stasis, causing longer postoperative hospital stay, this method has limited advantages over B-I anastomosis after distal gastrectomy.