The combination of alcohol and stress have been considered producers of gastric hemorrhage both experimentally and in clinical observations. Since excessive alcohol intake often occurs in situations of severe emotional conflict and stress, it was decided to study the possible role of the latter in the etiology of gastric hemorrhage, up to now thought to be dependent only on alcohol. The study consists of 75 male Wistar rats divided into eight groups with seven to 14 animals each. They were submitted to fasting only, or to additional prolonged fasting, restraint-stress (physical confinement) for 17 hours and the oral administration of a single dose of 40% alcohol (1 ml/150 g of body weight). The stomachs were analyzed macroscopically and microscopically for the presence of gastric hemorrhage, and the following was observed; 1) only 10% of the rats submitted to a 25 hour fast either isolated or associated with 17 hours of physical confinement, demonstrated gastric hemorrhage; 2) after an eight hour fast, the administration of alcohol to the rats either sacrificed immediately or maintained for 17 hours, revealed gastric lesions in only 33.3% and 28.5% respectively, without significant statistical difference between the two groups (P > 0.05); 3) administration of alcohol prior to the 17 hour physical confinement revealed lesions in only 12.5% of the animals; 4) administration of alcohol to rats previously submitted to a 25 hour fast plus physical confinement for 17 hours, resulted in a significant number of hemorrhagic lesions (88.8%). This caused a statistical difference in the group compared to the others (P < 0.01). The results of this study led to the conclusion that stress (by prolonged physical confinement) was an important conditioning factor to the appearance of gastric hemorrhage when 40% alcohol was administered. It is possible that if the alcohol had been administered prior to the prolonged physical confinement its cytotoxic effect on the gastric mucosa might have been reduced.