Background and Objectives: It is critical to distinguish benign from malignant thyroid nodule and to select a patient for surgery. Even though the U/S study dose not make great contribution to diagnose a malignant thyroid nodule, it is widely used in the evaluation of anatomic feature of thyroid. The authors tried to estimate the efficacy of the U/S study in preoperative diagnosis of malignant thyroid nodule. Materials and Method: At the department of General Surgery of Ajou University, 75 patients who were operated after diagnosis with thyroid nodule by U/S study between July 1996 to June 1997 were retrospectively analyzed. By comparing the U/S impression that implies malignant thyroid nodule to FNAC and post-operative pathologic results ware as follows. Results: 1) Absence of cystic change, presence of internal hypoechogenicity, lobulation, calcification, thick and irregular halo, and nodule more than 4cm in diameter on U/S were considered significant statistically for the diagnosis of malignancy(Chi-square test, p<0.05) 2) Presence of internal hypoechogenicity or thick and irregular halo has the validity in Logistic regression analysis. 3) FNAC was done in 65 case. 19 case were malignant, 11 case were suspicious and 46 patients were benign (sensitivity 52.6%, specificity 87%). 4) The findings of U/S which are hypoechogenic and thick and irregular halo show 82% sensitivity and 97% specificity. In combination with the findings of FNAC that imply benign or suspicious lesions, the sensitivity was 100% and the specificity was 97%. Conclusion: This study suggest that the hypoechogenicity and thick and irregular halo on U/S are important information for the diagnosis of thyroid malignancy which were considered benign or suspicious after FNAC.