Certain molecules, in particular steroids, have been observed to be outliers to quantitative structure-permeability relationships (QSPRs) for skin permeability (k(p)). Recently, however, many of the historical skin permeability data for these compounds have been found not to be consistent with more modern data. In this study QSPRs were re-analysed replacing the originally published steroid permeability data with those from more recent studies. A highly significant QSPR describing skin permeability in terms of the octanol-water partition coefficient (logP) and molecular weight (MW) was derived (log k(p)=0.74 logP-0.0091MW-2.39). This model is similar to those published previously. Statistical analysis of the residuals from the QSPR determined that the steroids are no longer outliers to this model. Thus, they may be considered to penetrate the skin by the same means as the majority of exogenous chemicals in this model.