Epidemiological studies, as well as animal models, indicate that iodine might be an immunogenic agent for the thyroid gland, at least in subjects predisposed to thyroid autoimmunity. This review presents data, either epidemiological or experimental, obtained in different conditions: constant and stable iodine status, either deficient, sufficient or excessive; long-term iodine prophylaxis; temporary supplementation with iodide (6-12 months) or iodised oil. Moreover, we also discuss data obtained in the general population, among subjects with euthyroid goiter, or autoimmune goiter, or even in women prone to post-partum thyroid diseases. It is concluded that the significant increase in the prevalence of autoimmune thyroid diseases in populations living in iodine sufficient areas should not prevent the implementation of the iodine prophylaxis.