Detection of thyroid cancer by thyroglobulin (Tg) assay in peripheral blood is useful in the absence of residual thyroid tissue, but it requires thyrotropin stimulation for maximal sensitivity and is affected by circulating antithyroglobulin antibodies. To avoid these drawbacks, thyroglobulin mRNA (Tg mRNA) assay in circulating blood has been proposed. Initial studies showed that Tg mRNA assay was more positive in patients with metastasis than in cured patients. Further studies showed controversial data. We measured Tg mRNA in 26 patients undergoing levothyroxine (LT(4)) suppressive therapy after total thyroidectomy for thyroid cancer and in 11 controls. The stage of the cancer was defined according to the findings of the latest whole-body (131)I scan and serum Tg performed under LT(4) withdrawal. Patients were classified as cured (negative scan, negative stimulated Tg, 8 patients), with metastasis (positive scan in extrathyroid bed regions, positive Tg, 7 patients), with thyroid remnants (positive scan in thyroid bed, positive Tg, 8 patients), and discordant cases (negative scan, positive Tg, 3 patients). RNA was extracted from blood and analyzed by quantitative reverse transcription-polymerase chain reaction (RT-PCR) using two sets of primers and internal probes specific for Tg mRNA. This method allowed the detection of Tg mRNA in thyroid biopsies. Tg mRNA was undetectable in all control subjects and in all patients with cured cancer, positive in 1 of 8 patients with thyroid remnants, and in only 1 of 7 patients with metastasis. In conclusion, our data do not support the usefulness of Tg mRNA measurements in blood for monitoring thyroid cancer.