The logarithm of the partition coefficient (log P) of low-molecular-weight organic compounds is a physicochemical parameter used extensively in structure-biological activity studies to model interactions of the compounds with nonpolar phases in vitro and in vivo. The partition coefficient can be determined between water and a number of nonpolar solvents. The most common nonpolar solvent is 1-octanol, but solvents such as benzene, carbon tetrachloride, and chloroform are frequently used as models for the nonpolar phases. The functional relationship between chemical structure and partitioning is not well-understood. In this paper, partition coefficient data for 50 solutes in six nonpolar solvent systems are analyzed by using principal components analysis. The objective of the work is to explore the relationship between solute structure and partitioning behavior for simple organic compounds. Two structural factors are found to be important, with the isotropic surface area being the most important. The isotropic surface area can be used to estimate log P in some solvents and as an independent variable in quantitative structure-activity relationships (QSAR). This is illustrated by estimating the rate of epidermal diffusion of steroids.