Well-differentiated thyroid cancer (WDTC) is the most frequent form of endocrine neoplasia. One of the main challenges in the management of this disease is distinguishing low-risk patients who can be treated by surgical resection of the lesion from those with a high likelihood of recurrence who need a more extensive approach, including total thyroidectomy and radioiodine ablation. A tissue microarray (TMA) comprising 410 cases of WDTC was constructed with risk estimates for the following features: extrathyroidal extension, lymph node metastases, and vascular invasion. The variables examined were morphologic classification, candidate genetic, and proteomic biomarkers. BRAF (Raf kinase type B) mutant carriers showed increased risk of developing invasion compared with wild-type (WT) cases. However, when classified morphologically, classic papillary thyroid carcinomas (PTC) showed much higher risk estimates for invasive features compared with follicular variant PTCs (FVPTC); within these morphologic subgroups, BRAF mutational status did not provide independent risk estimates. Staining intensities for membranous galectin-3 (Gal3), HBME-1, and CK19 and nuclear Gal3 were statistically validated as markers of aggressive behavior. Estrogen receptor beta (ERβ) was overexpressed in lesions with invasive behavior. The utility of these biomarkers remained statistically significant in the FVPTC. In contrast, a different set of biomarkers proved effective in classic PTC where upregulation of cyclin D1, loss of p27, and overexpression of ERβ were associated with invasive behavior. Different proteomic signatures validate the distinction of classic and FVPTC and provide a practical clinical mechanism to predict the thyroid cancer behavior and stratify patients for clinical management. ©2011 AACR.