The migration of ketoprofen through a series of simple gels that varied in solvent composition to simulate snapshots of a dynamically drying topical formulation was studied. Firstly, the release rate of ketoprofen was determined from formulations based on Cabosil and PEG 400, the proportion of which was varied to mimic progressively dryer states. Secondly, the apparent permeability of ketoprofen across the corresponding blank Cabosil gels was determined. Thirdly, the effect of macro viscosity on these data was probed by comparing permeation of ketoprofen across Cabosil and hydroxypropyl cellulose (HPC) gels of equal viscosity. Linear release profiles were produced for all formulations suggesting first-order release and the rate of ketoprofen liberated was inversely to the proportion of Cabosil, suggesting that the drier the film, the slower the rate of release. At the lowest level of thickener used (5%) the release rate was reduced to 45% of the control. At 25% the release rate was reduced to 24% of the control. The presence of the Cabosil had an even more dramatic effect on the apparent permeability of ketoprofen across the gels. At 5% Cabosil the apparent steady state flux was reduced to 4% of the control. At 25% the apparent steady state flux was reduced to < 1% of the control. Although the 0.5% HPC gel and the 1% Cabosil gel possessed identical macro viscosities, the permeation of ketoprofen through the HPC gel was almost double that of the Cabosil gel. The data from these experiments demonstrated that migration of active molecules through a gel is significantly affected by the amount of solvent present in, or lost from, the system. It is proposed that increased adsorption of active to the thickener plays a more important role than increased macro viscosity for reduced active release as the formulation becomes increasingly dry. Furthermore, such affects are profoundly influenced by the chemical nature of the thickener.