In developing countries such as Ethiopia, where chronic gastritis and peptic-ulcer disease are the most common endoscopic findings, it is important to study the association between Helicobacter pylori infection and gastroduodenal diseases. Both invasive and non-invasive diagnostic methods were therefore used to investigate 300, consecutive, adult patients with dyspepsia, from the gastrointestinal clinic of Tikur Anbassa University Hospital, Addis Ababa. The apparent overall prevalence of H. pylori infection varied according to the detection method employed. Culture revealed H. pylori in only 69% of the patients but this pathogen appeared more common when rapid urease tests (71%), PCR-denaturating gradient gel electrophoresis (91%), histopathology (81%), silver staining (75%) or stool-antigen tests (81%) were employed. Antibodies to H. pylori were detected, both by enzyme immuno-assay (EIA) and immunoblotting, in approximately 80% of the patients, whether the antigens used were of a reference strain or from a local isolate of H. pylori. When some of the EIA-positive and EIA-negative sera were cross-absorbed with antigens of Campylobacter jejuni and re-tested by EIA, the H. pylori-positive sera remained positive and the negative sera remained negative. Dyspeptic patients in Ethiopia, like most of those previously observed elsewhere in Africa, are often infected with H. pylori. It is important that the management of these patients should not be hampered by the misinterpretation of the African epidemiology of this pathogen.