Thyroid dysfunction and autoimmunity have been reported during type I interferon therapy, namely interferon-alpha for chronic hepatitis or interferon-beta for multiple sclerosis. To define the frequency of thyroid dysfunction and autoimmunity during interferon-beta treatment, 156 multiple sclerosis patients were prospectively followed up by 18 centers for 1 yr after starting interferon-beta-1b treatment. Serial clinical assessments and tests of thyroid and liver function and antithyroid autoantibodies (all performed by the same centralized laboratory) were conducted every 3 months. TSH and antithyroid autoantibodies against human TG or thyroid microsomal antigens were measured by immunoradiometric methods; free T3 and T4 were measured by chromatographic assays. Longitudinal occurrence of thyroid or liver alterations or of autoantibodies was analyzed with the generalized estimating equations method, correcting for the correlation of repeated measurements of the same subject over time. Pretreatment comparison with a control group of 437 healthy blood donors did not show significant differences in the frequency of thyroid dysfunction or antithyroid autoantibody positivity. During interferon-beta treatment, the de novo frequency of thyroid alteration was 8.3%, that of liver alteration was 37.5%, and that of antithyroid autoantibody was 4.5%. Generalized estimating equations analysis demonstrated that the frequency of liver alteration significantly increased during treatment compared with the baseline value (odds ratio, 7.03; confidence interval, 2.49-19.9), whereas that of thyroid alteration or of antithyroid autoantibodies did not. The frequency of thyroid dysfunction during interferon-beta treatment showed random, nonsignificant changes over time and, in addition, was not correlated to antithyroid autoantibody positivity.