The aim of this study was to determine the accuracy of initial endoscopy combined with histology and to define whether there is a point in following-up all gastric ulcers until complete healing. We have studied all patients with gastric ulcers documented at endoscopy during a 6-year period. Ulcers were macroscopically characterised as benign or suspicious for malignancy, and biopsies were taken. A follow-up endoscopy and histology was performed 4–6 weeks and 3 months after an anti-ulcer treatment. Resistant ulcers were treated surgically. All patients were followed-up clinically and endoscopically for a year after complete ulcer healing. 802 patients with gastric ulcers were enrolled. At initial endoscopy, 732 ulcers (91.3%) were macroscopically characterised as benign and 70 ulcers (8.7%) as suspicious for malignancy. In the group of endoscopically benign ulcers, only one (0.1%) had malignancy detected by biopsy in the first examination. None of these ulcers turned out to be malignant on subsequent examinations. From the suspicious for malignancy ulcers, 20 (28.6%) were proven to be malignant. Endoscopy may recognise with great accuracy benign ulcers, but it overestimates the malignant ones. The cost-benefit of serial follow-up endoscopies should be re-evaluated in ulcers that appear benign, and biopsies are negative at the initial examination.