Some patients with thyroid cancer develop a second primary cancer. Defining the characteristics of patients with double primary cancers (DPCs) is crucial and needs to be followed. In this study, we examine molecular profiles in DPC. We enrolled 71 patients who received thyroid cancer surgery, 26 with single thyroid cancer (STC), and 45 with DPC. A retrograde cohort was used to develop immunohistochemical expressions of mismatch repair (MMR) proteins and cell-cycle-related markers from tissue microarrays to produce an equation for predicting the occurrence of DPC. The multivariate logistic model of 67 randomly selected patients (24 with STC and 43 with DPC) identified that the expression of deficient MMR (dMMR) (odds ratio (OR), 10.34; 95% confidence interval (CI), 2.17–49.21) and pRb (OR, 62.71; 95% CI, 4.83–814.22) were significantly associated with a higher risk of DPC. In contrast, the expression of CDK4 (OR, 0.19; 95% CI, 0.04–0.99) and CDK6 (OR, 0.03; 95% CI, 0.002–0.44) was significantly associated with a lower risk of DPC. Collectively, dMMR, pRb, CDK4, and CDK6 have a sensitivity of 88.9% (95% CI, 75.1–95.8) and a specificity of 69.2% (95% CI, 48.1–84.9) for occurrence of DPC in all 71 patients. This is the first report to demonstrate the molecular differentiation of STC and DPC. Overall, the integral molecular profile performed excellent discrimination and denoted an exponential function to predict the probability of DPC.