3045 Background: Availability of targeted therapies in thyroid carcinoma (TC) has challenged the conventional treatment algorithms and established urgency for timely identification of targetable genetic abnormalities. Tissue-based next generation sequencing (NGS) is often limited by tumor insufficiency and slow turn-around time. Plasma-based circulating tumor DNA (ctDNA) NGS overcomes these barriers and has been widely adopted across advanced-stage solid tumors. To date, plasma-based NGS characterization of genomic alterations in TC has not been determined. Herein, we profile potential actionable mutations detected via ctDNA in patients with advanced TC subtypes. Methods: A retrospective analysis of Guardant Health, Inc database was performed using the commercially available Guardant360 plasma-NGS test on advanced metastatic TC samples collected between 2016 and 2021. Patients with papillary TC (PTC), follicular TC (FTC), poorly differentiated TC (PDTC), medullary TC (MTC), and anaplastic TC (ATC) were clustered into four groups (G1: ATC, G2: PTC, FTC, and PDTC, G3: MTC, and G4: unspecified TC). The landscape of genetic alterations, frequencies of alterations in clinically relevant genes, and tumor mutation burden (TMB) were analyzed. Results: Of the 1,108 patients included, 47.1% were male. The median age was 65 years old (range 13-98), and 0.18% (n = 5) patients were under 18 years old. Alteration frequencies of selected, clinically relevant genes are demonstrated in the table below. TMB analysis was performed on 315 samples, and the mean TMB was higher in G1 compared to G2, G3, and G4 (p = 0.0029, 0.0826, and 0.0112, respectively). Conclusions: Plasma-based comprehensive NGS by Guardant360 may be utilized in patients with advanced metastatic TC for detecting clinically relevant genetic alterations for the selection of available targeted therapies, immunotherapy, or determination of the clinical trial eligibility. Future validation of the clinical utility by analysis of paired tumor and plasma samples is warranted. [Table: see text]