Diffusion and sorption data of methanol and toluene in the ternary system methanol-toluene-poly(vinylacetate) (pvac) were measured by means of a magnetic suspension balance (MSB). Ternary diffusion coefficients were determined from sorption kinetics with a method of Crank [1975. The Mathematics of Diffusion, second ed. Claredon Press, Oxford] classically used for binary polymer solvent systems. Assumptions and limitations for this method are discussed. Sorption data of methanol and toluene in the ternary system were correlated with a modified Flory-Huggins theory taking four instead of three interaction parameters into account and compared with predictions of UNIFAC-FV. All four binary Flory-Huggins interaction parameters were correlated as a function of concentration, determined from binary polymer-solvent sorption measurements and from solvent-solvent equilibrium data. The modified Flory-Huggins approach for ternary systems predicts the two limiting cases for the binary polymer-solvent as well the two solvent-solvent equilibria. Diffusion coefficients determined from sorption kinetics measurements were correlated with free-volume theory predictions of Vrentas et al. [1984. Self diffusion in polymer-solvent-sol vent systems. Journal of Polymer Science, Polymer Physics Edition, 22, 459-469] and compared with a few available literature data from Surana et al. [1998. Diffusion and equilibrium measurements in ternary polymer-solvent-solvent systems using inverse gas chromatography. Industrial and Engineering Chemistry Research, 37, 3203-3207]. Cross diffusion terms are not taken into account. Diffusion and sorption data, determined with this gravimetric method, were used in model simulations with modified Flory-Huggins and free-volume theory to predict drying curves of ternary methanol-toluene-pvac solutions. The results were compared with previously published experimental results performed by means of Inverse-Micro-Raman-Spectroscopy (IMRS) from Schabel [2004. Trocknung von Polymerfilmen - Messung von Konzentrationsprofilen mit der Inversen-Mikro-Raman-Spektroskopie. Ph.D. Thesis, Universitat Karlsruhe (TH), Germany]. (c) 2007 Elsevier Ltd. All rights reserved.