The potential of immobilized artificial membrane (IAM) chromatography to estimate human oral absorption (%HOA) was investigated. For this purpose, retention indices on IAM stationary phases reported previously by our group or measured by other authors under similar conditions were used to model %HOA data, compiled from literature sources. Considering the pH gradient in gastrointestinal tract, the highest logkw(IAM) values were considered, obtained either at pH 7.4 or 5.5, defined as log kw(IAM)best. Non linear models were established upon introduction of additional parameters and after exclusion of drugs which are substrates either to efflux or uptake transporters. The best model included Abraham’s hydrogen-bond acidity parameter, molecular weight as well as the positively and negatively charged molecular fractions. For reasons of comparison between IAM chromatography and traditional lipophilicity, corresponding models were derived by replacing IAM retention factors with octanol–water distribution coefficients (logD). An overexpression of electrostatic interactions with phosphate anions was observed in the case of IAM retention as expressed by the negative contribution of the positively charged fraction F+. The same parameter is statistically significant also in the logD model, but with a positive sign, indicating the attraction of basic drugs in the negatively charged inner membrane. To validate the obtained models a blind test set of 22 structurally diverse drugs was used, whose log kw(IAM)best values were determined and analyzed in the present study under similar conditions.