The annual incidence of clinically diagnosed TC in Malmö was, on an average, 2.4 per 100,000 population during the years 1960-1977. This was 1.2 per 100,000 population lower than the corresponding incidence in the whole of Sweden as reported by the National Cancer Registry. The main reason for the difference was suggested to be inclusion in the official figures of autopsy cases and of cases with a benign diagnosis, rather than a true difference in the prevalence of TC. During the later part of the study an increase in the incidence of differentiated TC of approximately 70% was noted. This was considered to be due to increased health awareness and the availability of medical care, because only the number of tumors with less advanced growth increased. The average annual mortality from TC in Malmö was 0.9 per 100,000, which was 0.4 per 100,000 lower than the corresponding official rate in all of Sweden. The difference was suggested to be mainly due to inclusion in the official figures of persons not dying of TC. The mortality did not change significantly during the period of investigation. The percentage distribution by histologic type of tumors clinically diagnosed (N = 104) was as follows: papillary cancer, 65%; follicular, 21%; medullary, 4%; and anaplastic, 12%. The prognosis as estimated by the life table method was worst for patients with anaplastic TC, followed by those with follicular, papillary, and medullary TC. The validity of using the relationship of the tumor to the thyroid capsule (i.e., intrathyroidal and extrathyroidal growth) as a basis for classification into tumor stages was supported in the present study: the mortality in patients with intrathyroidal tumors was lower than in those with extrathyroidal tumors. The definition of occult TC--TC not larger than 1.5 cm, without regard to the relation to the thyroid capsule--was considered inappropriate and a change in the conception of occult TC was proposed. The presence or absence of node metastases in TC did not seem to have major significance for the prognosis. The significance of age for survival was strongly supported in our study. Deaths from TC clinically diagnosed before the age of 60 were infrequent, whereas the disease after this age increasingly often was fatal. This was partly due to a late onset of anaplastic TC and partly to a higher mortality in older than in younger patients with papillary or follicular TC.(ABSTRACT TRUNCATED AT 400 WORDS)