Diffusion in liquids can still be predicted only with high uncertainty due to the lack of sufficient experimental data. Diffusion experiments are complex and time-consuming. Furthermore, the concentration dependence of the diffusion coefficients requires usually several experiments even for binary mixtures. The possibility to extract this information from one short Raman diffusion experiment is explored here. A general identification framework is provided which does not require the a priori specification of a diffusion coefficient model structure but establishes the concentration dependence directly from the data. The methodology is used to determine the diffusion coefficient in the mixture ethyl acetate-cyclohexane in a wide concentration range. (c) 2004 Elsevier B.V. All rights reserved.