Background: There are currently 3 types of covered metal stents available in Europe for palliation of patients with malignant dysphagia. Their relative merits have not been compared in a prospective, randomized study. Methods: One hundred consecutive patients with esophagogastric carcinoma were randomized to placement of an Ultraflex stent, a Flamingo Wallstent, or a Gianturco-Z stent. Malignant strictures of the esophagus were treated by insertion of a small-diameter stent (n = 71), whereas those involving the gastric cardia were treated with a large-diameter stent (n = 29). Results: At 4 weeks, dysphagia had improved in all patient groups (p < 0.001), but the degree of improvement did not differ among the 3 groups (p = 0.14). There were differences among the 3 stent types with respect to major complications (Ultraflex stent: 8/34 [24%], Flamingo Wallstent: 6/33 [18%], and Gianturco-Z stent: 12/33 [36%]), but these were not statistically significant (p = 0.23). Nine patients (26%) with an Ultraflex stent, 11 (33%) with a Flamingo Wallstent, and 8 (24%) with a Gianturco-Z stent had recurrent dysphagia (p = 0.73), mainly because of tumor overgrowth or stent migration; 12 of 13 episodes of migration involved small-diameter stents in the esophagus. Conclusions: All 3 covered metal stents evaluated offer the same degree of palliation of patients with malignant dysphagia. Placement of Gianturco-Z stents was associated with more complications as compared with Ultraflex stents and Flamingo Wallstents. Although stent migration is reduced by increasing stent diameter, tumor overgrowth remains an intractable problem that requires a new approach. (Gastrointest Endosc 2001;54:145-53.)