Clinical aspects of 924 unselected patients with hyperthyroidism from 17 thyroid centres of 6 European countries were investigated in a prospective study for exactly 1 year. Serum samples were centrally assayed for thyroid hormones, thyroid autoantibodies and TSH-binding inhibiting immunoglobulins (TBII), and urine samples for iodine. 26 items of information per patient were computerized and evaluated. Forming 2 groups from iodine-deficient areas (IDA) and iodine-sufficient (ISA) according to the urinary iodine, it was possible to elucidate some characteristics independently of local factors. The most important findings were: 1. Three types of hyperthyroidism were described: Graves' disease was defined as hyperthyroidism with eye symptoms and/or the presence of measurable TBII; autonomous adenomas were defined by a single hot nodule in the thyroid scan; the remainder included TBII negative hyperthyroid patients as well as non-immunogenic forms of hyperthyroidism (toxic multinodular goiter and other multifocal autonomies). These were termed "non-classifiable" patients. Graves' patients, representing an average of 60% of the patients, have an age peak between 40 and 49 years; they are 10 years younger than unclassified hyperthyroid patients and 20 years younger than patients with autonomous adenoma, who represented 9.2% of the patients. Surprisingly, there was no difference in the prevalence of Graves' disease between IDA and ISA. 2. An unexpectedly high rate of 10.5% of hyperthyroid patients (Graves' disease 8.5%, non-Graves' disease 14.2%) had no goiter, a figure increasing in the elderly to almost 17%. In IDA the thyroid was larger and the goiter more often nodular than in ISA. IDA are characterized by significantly higher rates (11.3%) of autonomous adenoma than ISA (3.2%). 3. 59.4% of 507 patients with Graves' disease had eye symptoms, 44.6% of at least grade II or higher using the Werner classification. There was no difference between patients from ISA and IDA. We found unilateral ophthalmopathy in 6.7% of Graves' patients. 4. The median iodine excretion was calculated from urinary iodine after exclusion of values over 250 micrograms iodine/g creatinine, arbitrarily defined as iodine contamination. In IDA the median iodine excretion was 63.6, and in ISA 105 micrograms/g of creatinine. 5. Surprisingly, proven iodine contamination was as frequent in IDA (14.5%) as in ISA (15.2%). In the survey protocol, however, this was noticed less often by physicians in ISA.(ABSTRACT TRUNCATED AT 400 WORDS)