Seventy-five patients (50 M, 25 F), affected by chronic hepatitis caused by hepatitis C virus (HCV), without clinically overt thyroid disease, underwent r-interferon (IFN)-α-2a treatment (3-6 MU, 3 times/week) for 12 months. They were tested for thyroid function and for thyroid autoantibodies before (A), 6 (B) and 12 (C) months after the beginning of treatment and after 6 (D) months off therapy. Antithyroglobulin antibodies (Tg-Ab) and TSH were measured by IRMA, antiperoxidase antibodies (TPO-Ab), free T<sub>3</sub> (FT<sub>3</sub>) and free T<sub>4</sub> (FT<sub>4</sub>) by RIA, thyrotropin receptor antibodies (TR-Ab) by RRA. None of the patients showed TR-Ab positivity throughout the study. The number of the patients with one or both antithyroid antibodies progressively increased during treatment (A 10.7%; B 26.7%; C 45.3%) and decreased when off therapy (D 22.7%) with none of them positive for Tg-Ab alone (TPO-Ab 6.7%; Tg-Ab + TPO-Ab 16%). Tg-Ab increased during rIFN (median: A 29.0; B 35.0; C 73.0 U/ml) but decreased when off therapy (D 29.0 U/ml). Instead, TPO-Ab significantly increased throughout the study (A 1.0; B 3.0; C 6.0; D 7.0 U/ml). However, some patients showed for the first time an appearance of antibodies when off therapy. Five patients showed both antibodies and thyroid dysfunction: 2 at B, 2 at C, and 1 at D. Only 1 developed mild transient hyperthyroidism while the other 4 developed hypothyroidism, persistent however only in 1 case. Our study confirms that rIFN-α-2a frequently induces thyroid autoimmunity. TPO-Ab seems more useful than Tg-Ab in monitoring the immunological response. Thyroid dysfunction (especially hypothyroidism and in females), always accompanied with the appearance of thyroid autoantibodies, is often transitory in spite of not discontinued treatment. The pretreatment positivity for Tg-Ab was not associated with the subsequent development of thyroid dysfunction, therefore it cannot be considered a risk factor in IFN-α therapy. Finally, the appearance in some cases of TPO-Ab after stopping treatment suggests that the thyroid surveillance in these patients should be prolonged.