It has been suggested that profound acid inhibition by proton pump inhibitors affects the accuracy of H. pylori detection. This report aims to evaluate H. pylori status during treatment with four different invasive detection methods and to investigate if histopathological alterations during treatment can be used as an early marker for H. pylori eradication. Twenty-eight H. pylori-positive patients were studied randomized into two treatment groups: 14 patients received omeprazole, 20 mg plus amoxicillin 1,000 mg b.i.d (OA), and 14 patients received omeprazole, 20 mg and placebo b.i.d (OP) for 14 days. Biopsies from antrum and corpus of the stomach were collected on days 0, 10 and 42. H. pylori status was based on rapid urease test, cultivation, histology, and polymerase chain reaction (PCR). The biopsies were also graded according to the Sidney classification. In the OP and the OA group, 17% (2/12) and 92% (12/13) of the patients were H. pylori negative when tested during treatment (day 10). Four weeks after treatment none of the patients (0%) in the OP group and 61% (8/13) in the OA group had their H. pylori infection eradicated. PCR was up to 34% more sensitive than the other tests to detect H. pylori during treatment. There was a decrease in histological inflammation and activity in the antrum already during treatment in the OA group, but the decrease did not discriminate for successful treatment. During treatment with omeprazole alone or in combination with amoxicillin, H. pylori detection is impaired regardless of the detection method used. However, PCR appears to be more sensitive than other tests. Early changes in the histological appearance of the gastric mucosa do not predict H. pylori treatment outcome.