Immobilized artificial membrane (IAM) chromatography coupled to physicochemical descriptors was evaluated to model the passive intestinal absorption of drugs through rat gut sacs. The chromatographic capacity factors (logk'(IAM)) of 12 structurally diverse compounds were determined on a IAM PC DD2 column. The passive permeabilities (P(a)) of the drugs were determined through rat everted gut sacs or through non-everted sacs for actively transported molecules. Correlation studies between logk'(IAM), physicochemical descriptors and P(a) were conducted by stepwise multiple linear regression (MLR) and back-propagation neural network (BPNN). MLR and BPNN showed that logk'(IAM) was the descriptor which correlated best with P(a). Considering the molar volume as an additional descriptor, the correlation was improved. Retention indices on IAM and the molar volume can be used concurrently to predict passive drug absorption.