Thyroid cancer is highly common all around the world. Its prevalence has rapidly increased over the last 30 years in the United States and other developing countries. Fine-needle aspiration biopsy has become the cornerstone of thyroid nodule diagnosis, whose general reliability is outstanding; however, some aspirates have shown undetermined cytological findings that do not provide a definitive malignancy diagnosis. At least 70 molecular and genetic markers in thyroid nodules have been analyzed in an effort to identify molecular markers to differentiate malignant and benign (BN) thyroid nodules. The present review focused on the currently used markers in thyroid cancer diagnosis. A rising number of studies have investigated immunohistochemical markers, such as galectin-3 (GAL3), cytokeratin 19 (CK19), Hector Battifora Mesothelial-1 (HBME-1), and thyroid peroxidase, along with DNA alterations, including mainly BRAF (B-Raf proto-oncogene, serine/threonine kinase) and RAS (Ras proto-oncogene, GTPase) point mutations, Telomerase reverse transcriptase mutations, ret proto-oncogene/papillary thyroid carcinoma (RET/PTC) and PAX8/PPARG rearrangements, and miRNA signatures and circulating tumor cells for thyroid cancer diagnosis. Although certain markers are promising for differential diagnosis, due to limitations of the substantial prevalence of BN thyroid tumors, none of them is specifically definitive to a large extent. Herein, we also discussed the studies that have supported the use of combinations of several markers, like KIT, TC1, miR-222, and miR-146b combination, as well as GAL3, CK19, and HBME-1 combination, in enhancing the diagnostic accuracy in differentiating malignant and BN tumors.