Demonstration of Helicobacter pylori infection receives more and more importance in nowadays gastroenterological practice. The authors have compared culture and histology from 69 antral biopsy specimens for their ability to document Helicobacter pylori infection. Infection ratios in the context of clinical and histological diagnoses resulted in a distribution pattern similar to that described by others: 85-69% of duodenal ulcer patients, 67-67% of gastric ulcer patients, 62-54% of patients with gastritis and/or erosion(s) and 33-60% of endoscopically negative patients were found to be Helicobacter pylori positive with culture and histology respectively. Normal or atrophic mucosa showed no bacteria with either methods, but one must also consider the small number of such cases in this study. Chronic gastritis with no signs of activity proved to be infected only in a minority of cases, while chronic active gastritis cases were Helicobacter pylori positive in 72 and 61% histologically and with culture respectively. The modified Giemsa stain used in this study grave a relative specificity of 0.74 and sensitivity of 0.79 as compared with culture. There was a good association between the two methods tested, and this could be further improved by doing the two tests simultaneously. The association was weaker when the tests were done asynchronously, this is why it is not recommended to use such a diagnostic schedule. Comparison of the traditional haematoxylin-eosin stain with the modified Giemsa stain resulted in a very strong association between the two.(ABSTRACT TRUNCATED AT 250 WORDS)