The association between an indicator of iodine deficiency and differentiated thyroid cancer has been investigated in a population-based case-control study. This included all incident cases (N = 484) of papillary and follicular thyroid cancer diagnosed during the years 1980-1992 in a Swedish Health Care Region among residents born in Sweden, and as many individually matched controls. Cases were included after a uniform review of their histopathological specimens. Residence in areas where goiter had been severely endemic in the 1930s was used as exposure indicator. Odds ratios (OR) and 95% confidence intervals as estimates of relative risk were calculated as a measure of association using logistic regression. A trend toward an association was found with a duration of residence in goiter areas between 21 and 40 years, most prominent among follicular cancer cases and in the group diagnosed at age > or = 50 years compared to not exposed. Exposure for the first time during adolescence (between 11 and 20 years) was associated with an increased risk of papillary cancer. This was especially evident among women, both when compared to not exposed and to those exposed during the first year of life; the association was strengthened after adjustment for duration of residence. These findings support the hypothesis of distinct causation patterns from iodine deficiency to the 2 most common histological types of thyroid cancer.