Despite the considerable success in predicting the steady-state dermal absorption rates of chemical compounds from large reservoirs applied to skin, correspondingly little progress has been made in predicting the absorption rate and extent for small doses of topically applied compounds. In the latter case, steady-state absorption rates are generally not obtained, and rapid evaporation or penetration of the dose solvent makes application of permeability coefficient models problematic. This report presents a new analysis of the finite dose problem in terms of a diffusion model with three parameters-a characteristic time for diffusion, h2/D; a skin solubility factor, S(m)h; and a capacity factor for absorption of the dose during the dry down period, M*. These parameters can be related to the molecular weight and oil and water solubilities of the permeant in a manner similar to models describing steady-state absorption from saturated solutions. Some variation of the parameter values based on the chemical nature and volume of the dose solvent is anticipated. The applicability of the model is demonstrated by analyzing the in vitro absorption rates of varying doses of vanillylnonamide (VN, synthetic capsaicin) applied to excised human skin from propylene glycol. The analysis shows that a three-parameter model that assigns all of the resistance to transport to diffusion through the stratum corneum is able to explain most of the significant features of VN absorption through skin.