Having previously demonstrated that tissue miR-375 expression in medullary thyroid carcinoma (MTC) tissues is linked to prognosis, the aim of this study was to assess the diagnostic and prognostic value of circulating miR-375 levels in MTC patients. A series of 68 patients with MTC was retrospectively retrieved and assessed in terms of their clinicopathological characteristics. MiR-375 levels were measured in all patients' presurgical blood samples. Both serum and tissue levels were tested prior to surgery in a subgroup of 57 patients. Serum miR-375 levels were also measured in serum from 49 patients with non-C-cell thyroid nodular diseases (non-CTN), 14 patients with pheochromocytoma, and 19 healthy controls. Circulating miR-375 levels were 101 times higher in the serum of patients with MTC than in all other patients and controls, with no overlap (P < 0.01). No correlation emerged between serum and tissue miR-375 levels. Serum miR-375 levels were higher in MTC patients with N0 than in those with N1 disease (P = 0.01), and also in patients who were biochemically cured than in those who were not (P = 0.02). In the whole series of patients and controls, calcitonin (CT) and serum miR-375 levels were correlated at diagnosis (R = 0.40, P < 0.01), but in a U-shaped manner: a positive correlation was found with low CT levels, then the correlation turns negative as CT rises (in MTC patients). A negative correlation was indeed found in MTC patients between serum miR-375 and CT (R = -0.10, P = 0.01). On ROC curve analysis, a cut-off of 2.1 for serum miR-375 proved capable of distinguishing between MTC patients and the other patients and controls with a 92.6% sensitivity and a 97.6% specificity (AUC: 0.978, P < 0.01). Serum miR-375 levels can serve as a marker in the diagnosis of MTC, with a remarkable specificity. Serum miR-375 also proved a novel marker of prognosis in this disease. Further experiments to corroborate our results are currently underway.