Helicobacter pylori infection of the stomach is associated with gastritis and peptic ulceration and may be causative. A noninvasive test for this organism might be useful in managing some patients with dyspepsia without the need for further investigation. We have evaluated a new commercially available serological test (Helico-G ELISA, Porton Cambridge, UK) for this infection to assess its diagnostic accuracy in a retrospective study of 115 patients with non-ulcer dyspepsia. Sixty-three of these patients (55%) were found to have H. pylori infection and gastritis on histology. A sensitivity of 81% and specificity of 90% were obtained. No significant fall in the antibody titres was found in a subgroup of 15 patients who were selected to complete a course of triple therapy despite significant improvement in their dyspepsia score and confirmed eradication of H. pylori organism in 80% of these patients. We conclude that the test has limited value in aiding clinical decision of managing patients with dyspepsia.