BackgroundThe challenge of selecting thyroid nodules for fine-needle aspiration (FNA) cytology has led to the development of the Thyroid Imaging Reporting and Data System (TIRADS), primarily in two formats: EU-TIRADS and ACR-TIRADS. Clinical observations suggest imperfect risk assessment for TIRADS 3 nodules ≥20 mm. This study aimed to evaluate the efficacy of TIRADS systems in distinguishing benign from malignant nodules in this subgroup. Materials and MethodsFrom May 2023 to March 2024, 1,094 patients with thyroid nodules were referred for ultrasound (US) at a University Hospital. Data on clinical, ultrasound, cytological, and histopathological parameters were collected. Nodules ≥20 mm were categorized by EU-TIRADS and ACR-TIRADS, and their predictive performance for malignancy was assessed through post-thyroidectomy histopathology or FNA cytology (Bethesda classification). Results267 patients (mean age 60.3 ± 14.3 years; 46 men, 221 women) with 308 nodules were analyzed. Twenty-two malignancies and 286 benign nodules were recorded. Recalculating EU-TIRADS 3 performance using 25 mm and 30 mm thresholds (ACR-modified EU-TIRADS) avoided 24% and 41% of FNAs, respectively, while ACR-TIRADS would prevent 26.6% (p>0.05). Two malignancies were missed. ConclusionEU-TIRADS and ACR-TIRADS show similar efficacy when using a 25 mm FNA threshold. Raising the cut-off for FNA in EU-TIRADS 3 nodules could reduce unnecessary procedures but may increase the risk of missed malignancies, impacting patient outcomes.