A procedure has been developed for measuring the steady state rate of permeation of commercial solvents through living human skin. To get the most consistent results, it was necessary with some solvents to normalize the solvent permeation rate of a given skin sample with its [3H]water permeation rate. For other solvents this was not necessary, so the un-normalized data were used. High [3H]water permeation rate also was used as a criterion for "defective" skin samples that gave erroneous permeability rates, especially for solvents having slow permeability. The linearity of the steady state data was characterized by calculation of the "percent error of the slope." The following permeability rates (g/m2h) of single solvents were measured: dimethyl sulfoxide (DMSO), 176; N-methyl-2-pyrrolidone, 171; dimethyl acetamide, 107; methyl ethyl ketone, 53; methylene chloride, 24; [3H]water, 14.8; ethanol, 11.3; butyl acetate, 1.6; gamma-butyrolactone, 1.1; toluene, 0.8; propylene carbonate, 0.7; and sulfolane, 0.2. The effect of [3H]water saturation on the shape of the presteady state portion of the permeation curve was determined and found to be very dependent on the solvent. The permeability of mixtures of DMSO and octyl acetate were measured. No octyl acetate was detected and the permeability of DMSO was proportional to its mole fraction in the mixture. The effect of two hours of solvent exposure on the viability of skin (based on DNA synthesis) was measured and found to be very dependent on the solvent.