The permeabilities of many steroids through human skin have been previously measured and reported in the literature. Analysis of these data reveals that significant discrepancies exist between the measurements of Scheuplein et al. and those of other groups. Six of the 14 steroids which were examined by Scheuplein et al., aldosterone, corticosterone, estradiol, hydrocortisone, progesterone, and testosterone, have also been examined by other groups. For each of these steroids, the permeability measurements of Scheuplein et al. are lower than those reported by other groups by factors of between 5.0 and 77. Eight independent measurements of the permeability of estradiol are in good agreement with one another, but are greater than the value reported by Scheuplein et al. by factors of between 11 and 20. Several possible sources of experimental error, including the variability of the skin samples, the differences in the experimental temperature, the establishment of steady-state conditions, the use of radiolabeled drugs, and the skin preparation technique, have been considered and do not appear to account for the magnitude of the observed discrepancies nor for the fact that the data of Scheuplein et al. are consistently lower than those reported by other groups.