AIM: To explore how often a diagnosis of gastric neoplasia is made on routine, non-targeted biopsies taken for determination of Helicobacter pylori status, compared with directed biopsies from endoscopically abnormal mucosa. METHODS: Records of all patients with a biopsy diagnosis of gastric cancer or dysplasia during a two year period were reviewed to determine whether the biopsy had been targeted at an area of mucosal abnormality, and whether there was any evidence of dysplasia or malignancy before endoscopy. RESULTS: Of the 8907 endoscopic examinations that included biopsy, histology showed malignancy in 115 cases and dysplasia in 20. Of these, in 128 cases the biopsies were targeted from focal abnormal areas of mucosa, and six were from areas of diffuse mucosal thickening. In one case, adenocarcinoma was diagnosed in a patient with a "normal" endoscopic appearance; this patient was undergoing repeat endoscopy for previous dysplasia. CONCLUSIONS: Gastric malignancy or dysplasia was detected histologically in 1.5% of endoscopies that included biopsy. The performance of routine biopsies not targeted at a visible lesion from patients without previous diagnosis of neoplasia did not increase the detection of gastric malignancy. Such biopsies are indicated, however, if histological aspects of a patient's gastritis (such as atrophy or intestinal metaplasia) influence the clinical management, as in the treatment of helicobacter gastritis.