Hemigastrectomy for benign disease and Helicobacter pylori infection are risk conditions for the development of gastric cancer. Aim of the study was to compare gastric histology and precursor lesions of malignancy in these two conditions. The hemigastrectomy group included 351 consecutively endoscoped subjects operated for gastroduodenal benign disease. Six to ten biopsy specimens were routinely taken from the residual gastric mucosa. The intact stomach group included 2097 consecutively endoscoped symptomatic subjects, who did not receive eradication therapy against H. pylori. The histological findings were classified as normal mucosa (NM), chronic non atrophic gastritis (CNAG), chronic atrophic gastritis (CAG), intestinal metaplasia (IM) and dysplasia (DYS). One thousand and three intact stomachs were H. pylori negative, and 1094 showed H. pylori colonization. The age over fifty was a significant risk factor for the occurrence of IM (OR 2.52, P < or = 0.001) and DYS (OR 3.46, P < or = 0.001), while Hp-positivity was a risk factor for CNAG (OR 1.81, P < or = 0.001) and CAG (OR 3.88, P < or = 0.001). Gastroresection was associated to higher risk for CNAG (OR 1.53, P < or = 0.001) and DYS (OR 4.31, P < or = 0.001) and to a lower risk of CAG (OR 0.49, P < or = 0.001). Both in males and females the risk for CNAG was significantly higher in Hp-positive (males OR 1.92, P=0.000; females OR 1.70, P=0.000) and gastrectomized subjects (males OR 2.06, P=0.000; females OR 2.43, P=0.000). Gastrectomized males, furthermore, showed an increased risk for DYS (OR 5.82, P=0.000). The aged Hp-negative and Hp-positive subjects evidenced a significant risk for IM (respectively OR's 3.42, P=0.000 and 4.85, P=0.000); the risk for DYS was significant in aged Hp-negative subjects (OR 4.09 P < or = 0.020). The Hp-positive individuals evidenced a significant risk for metaplastic mucosal changes (OR 38.17, P=0.000). Subjects aged over forty at the time of surgery and those with a longer postoperative follow up endoscopy presented an increased risk for CNAG of the residual mucosa (respectively OR's 2.75, P=0.000 and 5.25, P=0.000). CNAG and IM were the most frequently observed mucosal lesions both in subjects operated for duodenal and gastric ulcer (respectively OR's 4.02, P=0.000 and 3.00, P=0.000). Our data support that hemigastrectomy for benign disease and H. pylori infection may induce an increased incidence for histological precursor lesions for gastric malignancy and suggest that carcinogenesis in a resected stomach may be different from that in the intact stomach.