Diffusion has been shown, in previous work, to be an important factor for human intestinal absorption. Literature data was used to create data sets of diffusion coefficients in water and in ethanol. Diffusion coefficients were available for 139 compounds in water and 59 compounds in ethanol. Diffusion coefficients in ethanol were shown to have an excellent correlation with Abraham descriptors for hydrogen bond acidity, hydrogen bond basicity and molecular volume, with an R-2 of 0.93 and an S.D. of 0.04. Equations were also developed for calculating aqueous diffusion coefficients. The best equation for diffusion in water contained all five Abraham descriptors and had an R-2 of 0.73 and an S.D. of 0.07. The models of diffusion were both validated by developing training set models and making predictions for a test set. Y-scrambling was also used to ensure that the models are better than expected from random correlations. Principal component analysis was used to verify that the equations are similar to existing Abraham equations of absorption, thus supporting the idea that diffusion may be used to predict the absorption component of the bioavailability of drugs. (C) 2011 Elsevier B.V. All rights reserved.