To evaluate the use of fine needle aspiration cytology (FNAC) in the management of benign and malignant breast disease in an outpatient breast clinic and to determine the guidelines for reporting on FNAC of breast lesions, we evaluated the test results in correlation with the histologic diagnosis. We evaluate the test results of 669 cases of FNAC of the breast in 17 months in our outpatient breast clinic in correlation with the histologic diagnosis, histologic subtype according to the Japan Mammary Cancer Society classification and mammography results in false negative cases. Among 669 cases, 25.3% were inadequate for cytologic diagnosis. An analysis of the results using the two-by-four contingency table, including cases with inadequate cytologic diagnosis, showed that 10.6% and 1.0% had false negative and false positive cytologic diagnoses, respectively; sensitivity and specificity were 76.9% and 91.6%, respectively. The false negative rate of 10.6% was higher than expected, and those belonged mostly to the "inadequate" cytologic diagnosis category even though they were discrete, obviously malignant masses clinically. To try to explain this, we compared the cytologic results and histologic subtype according to the Japan Mammary Cancer Society classification and the mammography results. Of 73 cases of infiltrating ductal carcinoma in which retrospective reexamination of the histology slide was possible, 27 (37.0%) were papillo-tubular type, 11 (15.1%) were solid-tubular type and 35 cases (47.9%) were scirrhous type. Although the scirrhous subtype was only 37.0% in the group in which a cytologic diagnosis of carcinoma was possible, more than two-thirds of the cases of false negative cytology were scirrhous type and had malignant mammographic findings. This finding was statistically significant. In cases of ductal carcinoma of scirrhous subtype, the FNAC tended to be inadequate and false negative, but mammography showed better discrimination in such cases.