The relationship between thyroid dysfunction and breast cancer (BC) is debated. To clarify this controversial issue, a prospective study on thyroid function in BC was performed. The prevalence of thyroid disease was examined in 102 consecutive BC patients with ductal infiltrating carcinoma after surgery and before starting any chemohormonal or x-ray therapy and in 100 age-matched control healthy women living in the same borderline iodine-sufficient geographic area. All subjects were submitted to clinical ultrasound thyroid evaluation and serum free T4, free T3, TSH, thyroperoxidase antibody, and thyroglobulin antibody determination. Fine needle aspiration was performed in all thyroid nodules. Estrogen and progesterone receptors (ER and PR, respectively) were assayed in 92 and 55 BC specimens, respectively. The overall prevalence of thyroid disease was 47 in 102 (46%) in BC patients and 14 in 100 (14%) in controls (P < 0.0001). The prevalence of nontoxic goiter was 27.4% in BC patients and 11% in controls (P = 0.003). Hashimoto's thyroiditis was found in 13.7% of BC patients and in only 2% of the controls (P < 0.005). Other thyroid disorders found in the BC group included 2 cases of Graves' disease, 2 of thyroid carcinoma, and 1 of subacute thyroiditis, whereas in the control group only 1 case of Graves' disease and none of the other disorders were found. Mean free T3, free T4, and TSH concentrations showed no difference between BC patients and controls. The prevalence of thyroperoxidase antibody was higher in BC patients than in controls (23.5% vs. 8%; P < 0.005), whereas the prevalence of thyroglobulin antibody was not different. In BC patients the presence of thyroid antibodies was more frequently associated with clinically detectable autoimmune thyroiditis (14 of 26, 51.8%; P = 0.03) and was more common in the younger group. The positivity of ER was found in 51 of 92 (55.43%) and that of PR was found in 26 of 55 (47.27%) BC specimens. No relationship was found among ER, PR status, and the presence of serum thyroid antibodies. In conclusion, 1) the present study provides evidence that the overall prevalence of thyroid disorders is increased in patients with breast cancer, and 2) thyroid autoimmune disorders, especially Hashimoto's thyroiditis, account to a large extent for the increased prevalence of thyroid disease in patients with breast cancer. This feature is independent from the ER and PR status of the primary tumor. The present findings call attention to the usefulness of screening for thyroid disease in any patient with breast cancer.