Objective To establish a next-generation sequencing panel for the molecular diagnosis of thyroid nodules. Methods The panel, named ThyNod Panel, was designed to detect single nucleotide variant(SNV), indel, fusion, copy number variant (CNV) and RNA expression levels in 112 thyroid nodules associated genes, including benign and malignant molecular markers, thyroid differentiation or function genes and cell identity marker genes. Results The ThyNod Panel was successfully constructed and applied, and 856 benign or malignant thyroid nodules were completed sequencing. Totally 676 (79.0%) thyroid nodules were detected mutations. In 627 thyroid nodules with definite pathological diagnosis, 17.6% were benign nodules and 82.4% were malignant ones, in which malignant pathological types included classical papillary carcinoma, follicular variant papillary carcinoma, and medullary carcinoma. The most common mutant genes detected were BRAF (n=426), followed by RET (n=68), RET/PTC fusion (n=68), DICER1 (n=35), telomerase reverse transcriptase (TERT) (n=35), HRAS (n=24), NRAS (n=23), neurotrophin receptor kinase （NTRK3） fusion (n=19), and eukaryotic translation initiation factor 1A, X-chromosomal (EIF1AX) (n=11). The malignant percentage was counted for the detected mutations, and the malignant related variants such as BRAF V600E and RET fusion were higher than 90.0%, while the overall malignant percentage of RAS-like variants was only 18.9%. Conclusions ThyNod Panel can efficiently identify genetic characteristics in thyroid nodules and be applied in the molecular diagnosis of thyroid nodules.