PURPOSE: This study was performed to confirm the differences in telomerase activities between breast cancer and benign breast disease and to evaluate the clinical significance of telomerase activity in the diagnosis of breast cancer by its application to cytologic specimens. METHODS: Thirty-six breast-cancer samples and 23 benign-breast-disease samples were analyzed to detect telomerase activity by using a modification of the telomeric repeat amplification protocol (TRAP). In 27 breast-cancer specimens, both the cytologic samples obtained by fine needle aspiration (FNA) and the tissue samples from the primary tumor were analyzed using the same method for comparison. RESULTS: Telomerase activity was detected in 7 of 23 (30.4%) benign breast diseases and 30 of 36 (83.3%) breast cancers and was found not to correlate with the prognostic or pathologic variables. In the diagnostic accuracy of breast cancer, the resultant sensitivity, specificity, and positive and negative predictive values for telomerase activity were 83.3%, 69.6%, 81.1%, and 72.7%, respectively. Telomerase assays of cytologic and tissue samples of 27 breast cancers detected telomerase activity in 20 of 27 (74.1%) cytologic specimens and in 23 of 27 (85.2%) tissue specimens. CONCLUSION: The results of this study suggest that the detection of telo merase activity in tissue or cytologic samples is useful in the diagnosis of breast cancer.