Objective. To determine the diagnostic value of transvaginal ultrasonography and hysteroscopy in patients with postmenopausal bleeding. Material and methods. Between January 1, 1998 and June 30, 1999, 88 outpatient women with postmenopausal bleeding were enrolled in a prospective study. They underwent transvaginal ultrasonography and hysteroscopy, and were submitted to directed biopsy during hysteroscopy. Findings were classified as normal endometrium, suggestive of atrophy, focal abnormality (benign or suspicious), and diffuse thickness (benign or suspicious). Data was compared with the final diagnosis, established by histological examination, as atrophy, benign pathology, atypical hyperplasia and endometrial carcinoma. Results. Among 88 women enrolled, 15 were excluded because hysteroscopy was impossible, and four had abandoned the study. The histological findings were scanty material in 12 (17.4%), atrophy in 24 (34.8%), cystic atrophy in one (1.4%), normal endometrium in five (7.2%), tuberculous endometritis in one (1.4%), polyps in 12 (17.4%), leiomyoma in one (1.4%), non-atypical hyperplasia in three (4.3%), atypical hyperplasia in one (1.4%) and endometrial carcinoma in nine cases (13.0%). For the assessment of endometrial carcinoma, ultrasonography revealed sensitivity 77.8%, specificity 93.3%, positive predictive value 63.6%, negative predictive value 96.6%; and hysteroscopy revealed sensitivity of 88.9%, specificity 98.3%, positive predictive value 88.9%, negative predictive value 98.3%. The combined use of both methods revealed sensitivity 100%, specificity 91.7%, positive predictive value 64.3%, negative predictive value 100%. Conclusions. Both imagiological methods were found to be useful screening tests for endometrial carcinoma. Hysteroscopy was a superior diagnostic procedure.