Although it is well known that radiation induces chromosomal aberrations, there is a lack of information on the in vivo dose-effect relationship in patients receiving iodine-131 treatment, and the results of previous studies are controversial. In this study, the sister chromatid exchange (SCE) method was employed to investigate acute and late chromosomal damage (CD) in the peripheral lymphocytes of 15 patients who received various doses of I-131 (259-3,700 MBq), either for thyrotoxicosis (TTX) or for ablation treatment in differentiated thyroid cancer (DTC). The SCE frequencies in cultured peripheral lymphocytes were determined before treatment (to assess basal SCE frequencies), on the 3rd day (to assess acute SCE frequencies) and 6 months later (to assess late SCE frequencies). The basal, acute and late SCE frequencies (mean+/-SD) were 3.19+/-0.93, 10.83+/-1.72 and 5.75+/-2.06, respectively, in the whole group, and these values differed significantly from each other (P<0.001). In order to perform a quantitative evaluation of the present data and a comparative analysis with the results of previous studies reported in the literature, we defined acute and late effects using a "damage ratio" (DR) and a "recovery ratio" (RR), based on the basal, acute and late data for individual patients. No statistically significant difference was found in the DR between DTC and TTX patients (76.4%+/-11.5% vs 67.6%+/-9.0%), while the mean RR was higher in TTX patients than in the DTC group (75.2%+/-24.4% vs 36.8%+/-13.7%). The DR on the 3rd day was not related to the administered I-131 dose in the whole group, but a negative correlation was found between the I-131 dose and the RR at the 6th month (r=-0.60, P=0.04). The best fit for this relationship was obtained by a linear-quadratic model, as y=104.89x-28.4x(2)+38.1 (R-2=0.51, P=0.04). On the other hand, comparative analysis with the results of previous studies with comparable sampling times revealed that the best fit for the relationships between the administered dose of I-131 and DR and RR were obtained with a linear-quadratic model (Y=alphaD+betaD(2)) rather than a linear one. However, there was an interesting difference in comparison with in vitro studies, in that we found the coefficient beta to have a negative value, suggesting the disappearance of damaged lymphocytes from the peripheral circulation in a dose-dependent manner following I-131 treatment. Further studies are therefore needed to clarify the effect of the negative beta value on the biological dosimetry approach in continuous internal low LET radiation, as in the case of I-131 treatment.