The interaction of urea and several naturally occurring protein stabilizing osmolytes, glycerol, sorbitol, glycine betaine, trimethylamine oxide (TMAO), and proline, with condensed arrays of a hydrophobically modified polysaccharide, hydroxypropylcellulose (HPC), has been inferred from the effect of these solutes on the forces acting between HPC polymers. Urea interacts only very weakly. The protein stabilizing osmolytes are strongly excluded. The observed energies indicate that the exclusion of the protein stabilizing osmolytes from protein hydrophobic side chains would add significantly to protein stability. The temperature dependence of exclusion indicates a significant enthalpy contribution to the interaction energy in contrast to expectations from 'molecular crowding' theories based on steric repulsion. The dependence of exclusion on the distance between HPC polymers rather indicates that perturbations of water structuring or hydration forces underlie exclusion.Solutes are widely used to modulate the stability of native or folded conformations of proteins and nucleic acids (1-7). There are several naturally occurring osmolytes that cells synthesize to protect proteins in response to denaturing environmental conditions such as heat shock. Stabilization of compact structures typically results from an increased exclusion of solutes from the unfolded or more open conformations. There is an unfavorable interaction of solutes with exposed surfaces. The exclusion of osmolytes from surfaces necessarily means the inclusion of water and has quite naturally been termed a preferential hydration (6). Excluded, stabilizing osmolytes that are naturally occurring include glycerol, sorbitol, glycine betaine, proline, and trimethylamine oxide (TMAO) (8). Denaturation results when there are favorable interactions of solutes with exposed surface; more solutes are 'bound' or included with unfolded structures. Urea is probably the best known denaturant. The nature of the interaction between the solute and the macromolecule that results in exclusion or inclusion has not been satisfactorily characterized. Crowding theories that have been successful for the interaction of macromolecules (9) have been reformulated for small solute-macromolecule forces (10). This, however, does not explain the chemical specificity of the interaction. Bolen and coworkers, for example, have concluded (8,11,12) that the inclusion of urea and the exclusion of stabilizing osmolytes from proteins are dominated by the interaction of these small molecules with the peptide backbone with little contribution from the exclusion of these polar solutes from hydrophobic side chains. One method for elucidating the physics of the interaction is to measure the distance dependence of the force through either a radial distribution function of solutes †