Although commonly practiced in Japan, the effectiveness of regular screening with upper gastrointestinal (UGI) endoscopy against gastric cancer has not been well evidenced. The aim of the study was to investigate if gastric cancer-related mortality can be reduced by regular endoscopy. The medical records of 833 patients with gastric ulcer (GU) and 2547 without ulcer (NU) were analyzed; these patients received long-term, repeated endoscopic examinations between 1969 and 2004. Gastric cancer incidence, death by gastric cancer, and overall survival were compared with those in a Japanese general population by calculating the standardized incidence ratios (SIRs) and standardized mortality ratios (SMRs). The interval between UGI endoscopic examinations was 1.4 +/- 1.4 years in the GU group and 1.8 +/- 1.5 years in the NU group. During follow-up, 32 patients with GU and 61 in the NU group developed gastric cancer, showing annual incidence rates of 0.40% (95% CI: 0.24-0.56%) and 0.38% (0.28-0.48%), and SIRs of 2.21 (1.44-2.98) and 1.72 (1.29-2.15), respectively. The 5-year survival rate exceeded 80% among patients who developed gastric cancer. SMRs for gastric cancer and overall deaths were 0.50 (0.01-0.99) and 1.05 (0.87-1.23) in GU patients, and 0.45 (0.15-0.74) and 0.78 (0.69-0.88) in NU patients. There were no significant differences between the two groups in gastric cancer incidence, mortality from gastric cancer, and overall survival. Mortality from gastric cancer could be reduced by regular UGI endoscopy in a population with a high incidence of gastric cancer.