Vapor pressure data (at 50 degrees C) of solutions of poly(methyl methacrylate) [PMMA], polystyrene [PS], and poly(styrene-ran-methyl methacrylate) [P(S-ran-MMA)], with different weight fractions f of styrene units, in either CHCl3, acetone [AC], methyl acetate [MeAc], or toluene [TL] were evaluated with respect to the dependence of the Flory-Huggins interaction parameter X on polymer concentration and on f. For all solutions under investigation, X varies considerably with the composition of the mixture, and only for four of them [CHCl3/PS, AC/PMMA, MeAc/PS, and TL/P(S-ran-MMA) f = 0.5] is this dependence linear; another four systems exhibit a minimum [CHCl3/PMMA, CHCl3/P(S-ran-MMA) f = 0.5, TL/PMMA, and TUPS], and only one [MeAc/PMMA] shows a maximum. With the exception of CHCl3/P(S-ran-MMA) and f = 0.5, the chi values of the copolymers do not fall reasonably between the data obtained for the corresponding homopolymers. In most cases, the incorporation of a small fraction of the monomer that interacts less favorably with a given solvent suffices to make the copolymer behave very similar to the worse soluble homopolymer. In order to rationalize these multifaceted findings, we have modeled the results by means of an expression for chi, which accounts explicitly for the chain connectivity of polymers and for the capability of the components to change their molecular conformation upon mixing. This approach yields a consistent picture of the present results; it reproduces the effects of polymer concentration On v quantitatively by means of three physically meaningful parameters. The modeling of the influences of copolymer composition requires one additional term for each of these parameters.