To study the association between smoking and thyroid disease (Graves' disease [E05.0], nodular toxic goiter [E05.2], and autoimmune hypothyroidism [E03.9]) in a low-iodine intake area a case-control study was undertaken. A self-administered questionnaire was issued to 864 consecutive patients with hyperthyroidism and 628 patients with autoimmune hypothyroidism treated at five university or regional endocrinologic clinics in Denmark between January 1, 1990 and December 31, 1998. Each respondent was compared to an age- (+/- 5 years) and gender-matched normal control person randomly drawn from the background population. A total of 621 patients with hyperthyroidism (72%) and 411 patients with autoimmune hypothyroidism (66%) responded. Of these, 617 (542 females) and 408 (364 females) could be analyzed, respectively. There was an increased risk of both Graves' disease (odds ratio [OR] = 2.5, 95% confidence interval [CI]: 1.8-3.5), toxic nodular goiter (OR = 1.7, 95% CI: 1.1-2.5), and autoimmune hypothyroidism (OR = 1.5, 95% CI: 1.1-2.1) with ever smoking compared to never smoking in women, but not in men. With the high proportion of ever-smokers among women (56%), the attributable risk of smoking in women was 45% in Graves' disease, 28% in toxic nodular goiter, and 23% in autoimmune hypothyroidism. Ever use of oral contraceptives was associated with a slightly lower risk of Graves' disease in women, but not of toxic nodular goiter or autoimmune hypothyroidism. In conclusion, smoking is a powerful risk factor for thyroid disease, especially in populations with a high smoking frequency. Oral contraceptive use is associated with a slightly lower frequency of Graves' disease.