The practice of following benign-appearing gastric ulcers until healing was critically evaluated in a retrospective manner by reviewing all gastric ulcers that were followed with serial endoscopy and all gastric cancers diagnosed at the University of Alabama at Birmingham. The stated purpose of following ulcers to healing is to detect those gastric cancers that may be masquerading as benign ulcer and were not correctly diagnosed at initial endoscopy. Over a five-year period, 148 gastric ulcers were followed with serial endoscopy and in no case was an unsuspected carcinoma found at follow-up endoscopy. In addition, of 67 gastric cancers diagnosed between 1979 and 1986, 62 were suspected of being malignant by the endoscopist at initial examination for an accuracy of 92%. The accuracy rate based solely on biopsy and/or brush cytology was 94%. When endoscopic and biopsy and/or cytology impressions were combined, only one case of gastric carcinoma was not suspected. The overall accuracy was 99%. These results suggest that if either the endoscopic impression or the biopsy and cytology is suspicious for malignancy, then follow-up endoscopy until healing should be done. On the other hand, if, at the initial examination, the ulcer appears benign and biopsy plus cytology are negative, then serial endoscopy has a low benefit relative to its cost.