The induction of tumors by ionizing radiation is probably one of the best documented of all iatrogenic hazards. Nevertheless, quantitative data relative to the carcinogenic potential of radiation exposure in man are rare. This paper attempts to analyze the observed latent intervals between radiation exposure of the thyroid and the subsequent diagnosis of thyroid cancer, and to examine the distribution of these intervals with respect to age and sex. Material The authors have collected 528 cases of thyroid cancer in which a history of previous irradiation, with the thyroid gland probably included in the beam, was obtained. There is at present no way in which it can be proved that any one of these tumors was radiation-induced. As discussed elsewhere (1–6), however, strong circumstantial evidence exists in two forms. Compared to suitable controls: (a) groups of thyroid cancer patients have a substantially higher incidence of irradiation in their histories and (b) groups of patients who have received radiation therapy have a substantially higher incidence of thyroid cancer. For the purposes of the present study, it was assumed that radiation was at least a contributing cause in all the cancers reported here. Two hundred and nine cases are from the series published by Winship and Rosvoll in 1961 (7).4 Additional ones collected from the literature include a total of 57 from 15 sources listed previously (6) plus 147 from 15 more recent publications (8–22).5 Care was taken to avoid double registration of those cases which appeared in more than one series. Individual physicians contributed 80 unpublished cases.6 The files of the Hospital of the University of Pennsylvania yielded another 35 not previously published. The stated indications for irradiation are shown in Table I. For the most part, specific indications were associated with specific age groups. Nearly all patients who received radiation therapy because of a diagnosis of thymic disease were infants. Those treated for tonsil and adenoid disease, or cervical adenitis, were mostly older children. The acne patients were adolescent. Twelve of the 15 women treated for thyrotoxicosis and 10 of the 12 goiter patients were over twenty-one years old. A few patients received more than one series of treatments; for these only the first exposure was scored. Results A. Distribution of the Latent Intervals: The frequency distribution of the latent intervals is shown in Figure 1. All intervals were recorded as the nearest whole number of years. The ratio of females to males is 1.9, and is fairly constant over the full range. The distribution is seen to skew to the right. The mean latent interval, twelve and a half years, is to right of center in the distribution and is exceeded by only 38 per cent of the cases. The median, which represents the time at which approximately half the cases have appeared, is ten years. The shape of Figure 1 suggests a log-normal distribution.