A mathematical model based on mass transfer in plant tissues was developed with the aim of finding a simple predictive model which was easy to use and yet had a broad application scope in the osmotic dehydration of foods. The model takes into account the capacity of each constituent to diffuse within tissue by using a diffusion coefficient and a transmembrane mass transfer coefficient. The model also depends on the mass ratio of osmotic solution to product, on the initial chemical composition of the product and solution, and on the product shape. The application of the model to apple dehydration with polyethyleneglycol (PEG) solutions allowed us to quantitatively simulate the time evolution of cellular and extracellular volumes which was observed previously in histological sections under the microscope. By using experimental data obtained during potato dehydration in mixed solutions of sucrose and salt, mass transfer and apparent diffusion coefficients were fitted in order to evaluate product behavior under different process conditions. A good quantitative agreement was obtained between the experimental and calculated results. The model presented in this paper allows the main variables of osmotic dehydration to be estimated using a low number of fitting parameters, and interprets, through a simple algorithm, the complex phenomena of the main mass transfer mechanisms in plant tissues.