Data on transfer of radioiodine into human milk are rare in the literature. Data from sixteen publications were reviewed and analyzed to estimate the transfer coefficient (f(hm)*, having units of d L(-1)). The data on the radioiodine concentration in breast milk were analyzed by two methods: direct numerical integration and integration of a fitted exponential model. In general, the integrated fitted functions were greater. The fitted functions likely better describe the transfer into milk since few data sets sampled mothers' milk near the time of maximum excretion. The derived transfer coefficient values seem to represent two populations. The first group was those individuals who had very low excretions, including those where thyroid and mammary uptake was impaired by the administration of stable iodine or iodinated compounds. The second group included those with much higher excretions. The second group, termed the "normal-excretion" group, had transfers of iodine to milk that were more than ten-fold higher than in the "low-excretion" group. The derived milk transfer coefficient data for the low- and normal-excretion groups fitted to lognormal distributions gave geometric means, (geometric standard deviations), of 0.043 d L(-1) (2.1, n = 14) and 0.37 d L(-1) (1.5, n = 12), respectively. Estimates of the effective half-time (time from maximum concentration to half the value) were determined for the low- and normal-excretion groups separately. There was evidence that the effective half-time was longer for the normal- than for the low-excretion group; the geometric mean (and geometric standard deviation) were 12 (1.7) and 8.5 (2.6) h, respectively, though the difference was not statistically significant. The geometric mean times to maximum milk concentration in the low- and normal-excretion groups were nearly identical, 9.4 (3.1) and 9.0 (1.6) h, respectively. The data show that administration of large doses of stable iodine (commonly used to block uptake of iodine into the thyroid) is also an effective means to block radioiodine transfer into milk. Thus, protecting the mother's thyroid also protects the nursing infant. Despite inadequacies of available data describing the transfer of radioiodine to human milk within a healthy population of women, the values of f(hm)* provided here are believed to be the best available for use in radiological assessments. These values are particularly applicable to lactating women having normal diets and availability to stable iodine, as in the United States.