Radiolabeled compounds with varying partition coefficients (paraoxon, benzoic acid, parathion, and DDT) were chosen to study the percutaneous penetration and extent of dermal retention in pig skin both in vitro and in vivo. Radiolabel distributions within the skin were determined from 1 min to 24 h after application in ethanol. The distribution of radioactivity in the skin during the first 4 h was comparable between in vitro and in vivo experiments. At 24 h, radioactive residues in the dermis were significantly higher in vitro than in vivo for DDT, the most lipophilic compound. Increasing air flow over the skin surface significantly increased evaporative loss for volatile compounds (benzoic acid, N,N diethyl-m-toluamide, malathion, parathion, and DDT), significantly decreased the residues in the upper skin layer for N,N-diethyl-m-toluamide, malathion, parathion, and DDT, significantly decreased the dermal residue for malathion, and significantly decreased the penetration of N,N-diethyl-m-toluamide, malathion, and parathion. On a percentage basis, increasing the dose of parathion and paraoxon from 4 to 1000 mu-g/cm2 resulted in significantly lower residues in the dermis. When applied to the dermis, the more hydrophilic benzoic acid and paraoxon better penetrated the dermis than the more hydrophobic parathion and DDT. An ethanol vehicle facilitated the penetration of parathion into the dermis and receptor fluid. These results indicate that the dermis interacted with the penetrant during both in vitro and in vivo percutaneous absorption. Factors such as partition coefficient and dose of the penetrant, air flow over the skin, and vehicle changed the distribution of penetrants in the skin and percutaneous penetration. The dermis may exhibit a depot for lipophilic compounds during in vitro percutaneous absorption that is not observed in vivo. Formation of this depot was related to diffusibility of penetrant through the dermis rather than limited solubility in the receptor fluid.