OBJECTIVE To investigate the effect of varying amounts of iodine intake on the prevalence of thyroid dysfunction, autoimmunity and goitre in old age. METHODS The first screening study where elderly subjects with varying amounts of iodine supply but from the same geographical and ethnographical region (Carpathian basin) were compared, and all hormone measurements and ultrasonography were performed by the same laboratory or person. METHODS Nursing home residents were screened for thyroid disorders from: (A) an iodine-deficient area, Northern Hungary (n = 119; median age 81 years; median iodine excretion (MIE) 0.065 mumol/mmol creatinine (equivalent to 72 micrograms/g creatinine); (B) an area of obligatory iodinated salt prophylaxis since the 1950s, Slovakia (n = 135; median age 81 years, MIE 0.090 mumol/mmol creatinine (equivalent to 100 micrograms/g creatinine)) and (C) an abundant iodine intake area, Eastern Hungary (n = 92; median age 78 years; MIE 0.462 mumol/mmol creatinine (equivalent to 513 micrograms/g creatinine)). METHODS TSH, T4, free T4, T3, thyroglobulin (Tg), antibodies to Tg (AbTg) and to thyroid peroxidase (AbTPO), iodine excretion, ultrasonography of the thyroid gland. RESULTS In regions A, B, and C, the prevalence of unsuspected clinical hypothyroidism was 0.8%, 1.5% and 7.6% (P = 0.006), with all cases except one being antibody positive (Ab+). The occurrence of subclinical hypothyroidism was 4.2% in region A, 10.4% in region B and 23.9% in region C (P < 0.001), but only 3 of 22 cases with subclinical hypothyroidism from region C were Ab+. The overall prevalence of Ab positivity (either antiTg+ or antiTPO+) was similar in the three regions (A, 19.3%; B, 24.4%; C, 22.8%). The occurrence of hyperthyroidism (clinical plus subclinical) was 3.4% in region A, 3.0% in region B and 0% in region C (not significant). The rate of elevated Tg levels was similar in the three regions. The prevalence of goitre was 39.4%, 16.4% and 12.2% (P < 0.001), respectively in regions A, B and C. In euthyroid subjects the mean ultrasonographically determined thyroid volume was 21.9 ml in region A, 13.6 ml in region B and 15.1 ml in region C (ANOVA F = 5.76; P = 0.0038). There was no significant difference in the occurrence of cases with hypoechogenic echotexture of the thyroid gland. CONCLUSIONS The screening for hypothyroidism in nursing home residents living in iodine-rich regions is justified by the high prevalence of unsuspected clinical hypothyroidism. The high prevalence of antibody positivity in old age is independent of the iodine supply, but iodine supply has a determining role in the development of autoimmune hypothyroidism in the aged. Most cases of subclinical hypothyroidism in iodine-rich regions are not of autoimmune origin. In old age, hypoechogenic texture of the thyroid gland is not predictive of thyroid dysfunction.