Objective: Iodine is essential for normal thyroid function and the majority of individuals tolerate a wide range of dietary levels. However, a subset of individuals, on exposure to iodine, develop thyroid dysfunction. In this double-blind trial, we evaluated the efficacy and tolerability of low-dose iodine compared with those of levo-thyroxine (T 4 ) in patients with endemic goitre. Methods: Sixty-two patients were assigned randomly to groups to receive iodine (0.5 mg/day) or T 4 (0.125 mg/day) for 6 months. Subsequently, both groups were subject to placebo for another 6 months. Thyroid sonography, determination of thyroid-related hormones and antibodies, and urinary excretion of iodine were carried out at baseline and at 1, 6 and 12 months. Results: At 6 months, markedly increased urinary values of iodine were found in patients receiving iodine (36 mg/24 h at baseline, 415 mg/24 h at 6 months) compared with those receiving T 4 (47 mg/ 24 h at baseline, 165 mg/24 h at 6 months; P < 0.0001 compared with iodine group). T 4 administration engendered a greater (P < 0.01) decrease in thyroid volume (from 32 ml to 17 ml, P < 0.0001) than did intake of iodine (33 ml to 21 ml, P < 0.005). High microsomal and thyroglobulin autoantibody titres were present in six of 31 patients (19%) receiving iodine, and iodine-induced hypo-and hyperthyroidism developed in four and two of them, respectively. Fine-needle biopsy revealed marked lymphocyte infiltration in all six. After withdrawal of iodine, thyroid dysfunction remitted spontaneously and antibody titres and lymphocyte infiltration decreased markedly. Follow-up of these six patients for an additional 3 years showed normalisation of antibody titres in four of them. Conclusion: Although nearly comparable results were obtained with both treatment regimens regarding thyroid size, partly reversible iodine-induced thyroid dysfunction and autoimmunity were observed among patients with endemic goitre.