To compare the efficacy of imprint cytology, histology and CLO-test (for biopsy urease) in detecting Helicobacter pylori infection, antral biopsies were taken from 239 patients undergoing upper gastrointestinal endoscopy. Both imprint cytology and histology showed the presence of H. pylori in 215 (90%) patients. The sensitivity and specificity of imprint cytology vis-à-vis histology was noted to be 100%. The CLO-test was performed in 165 patients and was positive in 130 (79%) patients. The sensitivity and specificity of the CLO-test were 89% and 95%, respectively. The median time required for the CLO-test to become positive and for imprint was 60 minutes for each. The sensitivity of the CLO-test was reduced further in patients receiving colloidal bismuth subcitrate. Of the 27 patients receiving the drug the sensitivity of the CLO-test was only 9% after 4 weeks of therapy. However, the specificity was 100%. The sensitivity and specificity of imprint cytology were unaffected by the antimicrobial therapy and after 4 weeks of treatment were still 100%. It is concluded that the CLO-test has a lower sensitivity and specificity for diagnosing H. pylori infection compared to imprint cytology, which had a sensitivity and specificity equal to that of histology. Imprint cytology may be prepared as an adjunct to histology in patients in whom antral biopsies are taken as it offers a relatively quick diagnosis of H. pylori infection, is considerably cheaper than the CLO-test and does not require additional biopsy material.