Background: Despite the rising incidence of thyroid incidentalomas, their clinical significance remains unclear. The present study aimed to determine whether fluorodeoxyglucose positron emission tomography (FDG-PET) is associated with a significantly higher risk of clinically-significant thyroid carcinoma (CSC) in incidentalomas than other non-functional imaging modalities. Methods: Over a 2-year period, 89 patients were identified as having a thyroid incidentaloma. All patients had either surgery or fine needle aspiration cytology (FNAC) with a 12-month follow up to confirm the nature of the incidentaloma. Surgery was carried out for nodules with malignant or indeterminate FNAC result, or those with in a retrosternal location, with size > 4 cm or local symptoms. Results: A total of 21 (23.6%) patients had their incidentaloma detected by FDG-PET (PET group) and 68 (76.4%) by non-PET imaging modalities (non-PET group). Differentiated thyroid carcinoma was confirmed in 18 (20.2%) patients. The rate of malignancy was 61.9% in the PET group and 7.4% in the non-PET group (P = 0.001). After excluding the occult microcarcinomas, the risk of malignancy reduced to 14.6%, but the difference in malignancy rate became more marked between the PET and non-PET group (42.9% vs 2.9%, P = 0.001). The maximum standardized uptake value on FDG-PET was similar between benign and malignant lesions (P = 0.124). Conclusion: The overall risk of CSC in thyroid incidentalomas was 14.6%. Those detected by FDG-PET were significantly more likely to harbour CSC than those by non-functional modalities. Incidentalomas with focal FDG uptake should be thoroughly investigated with USG and FNAC.