BACKGROUND: Ultrasound (US) risk stratification systems (RSSs) have been developed to reduce the number of unnecessary fine needle aspirations (FNAs) of thyroid nodules. These systems were designed primarily to identify papillary thyroid carcinomas, thus their performance on follicular thyroid carcinoma (FTC) is debatable. The present study was undertaken to investigate the accuracy of RSSs in selecting FTCs for FNA. METHODS: Patients with FTC who underwent US examinations between 2012 and 2018 in 2 institutions were selected. US images were reviewed retrospectively, and FTCs were reclassified according and Thyroid Imaging Reporting and Data System (TIRADS). Risk class and indication for FNA were assessed. RESULTS:Forty-five FTCs from 45 consecutive patients were included in the study. The median tumor diameter was 32 mm (range, 11-100), and ovoid isoechoic nodule with or without lobulated margins was the most frequent presentation. When FTCs were classified according to RSSs, the most common categories were intermediate and high risk, though 1 case in 3 was not classifiable. FTCs were classified as high risk/high suspicion/malignant in 11% to 74% of cases, with a statistically significant difference among the systems. FNA was indicated in 69% to 100% of cases, with good agreement among AACE/ACE/AME, ACR-TIRADS, and TIRADS. CONCLUSION: Current RSSs show high performance in selecting FTCs for FNA. This result is mainly due to the dimensional RSSs cutoffs indicating FNA. On the contrary, given the reported unsuspicious echo-structural presentation of FTC and the recognized limitation of cytological assessment to detect it, caution is advised when using US to manage cytologically indeterminate nodules. Cancer Cytopathol 2020;128:250-259.