The present study aims to compare the diagnostic performance of two ultrasound classifications of thyroid nodules: 2015 American Thyroid Association (ATA) guidelines and 2017 American College of Radiology (ACR) Thyroid Imaging, Reporting and Data System (TI-RADS). This comparison evaluates the association of these classifications with the Bethesda system regarding the malignancy of thyroid nodules. A total of 166 patients who met the study's inclusion criteria were enrolled. Ultrasound examinations of the thyroid nodules were conducted, along with fine-needle aspiration (FNA) from the thyroid nodules, with subsequent submission to pathology. The mean age of the patients was 49.79 ± 15.61 years, and 84.3% of participants were female. The Bethesda category 2, with a frequency of 55.42%, was the most observed. 78.3% had solid nodules, with 48.2% showing isoechogenicity. The average nodule size was 22.56 ± 11.05 mm. The most frequent ATA and TI-RADS classifications were low (50.6% and 47.6%, respectively). Significant correlations were found between the Bethesda and TI-RADS scores (p < 0.001), the Bethesda and ATA scores (p < 0.001), and the ATA and TI-RADS scores (p < 0.001). The sensitivity of ATA and TI-RADS was 43% and 47%, respectively, with specificities of 98% and 98%. Both classification systems represent valuable tools for assessing thyroid nodule malignancy.