Non-bioadhesive, fully biodegradable soluble polymers would be very instrumental in advanced biomedical applications, such as gene and drug delivery and tissue engineering. However, rational development of such materials is hindered by the complexity of macromolecule interactions with biological milieu. The prevalence of carbohydrates in naturally occurring interface structures suggests an alternative, biomimetic approach. Interface carbohydrates, regardless of their biological function, have common non-signaling substructures (e.g., acetal and ketal groups, secondary and primary alcohols). We hypothesized that hydrophilic polymers (polyals) consisting of acyclic units built of non-signaling carbohydrate substructures would be highly biocompatible and non-bioadhesive, while intrachain acetal or ketal groups would enable nonenzymatic biodegradation upon uptake by cells. Acyclic hydrophilic polyals can be prepared via either polymerization of suitable monomers or lateral cleavage of cyclic polyals (e.g., polysaccharides). In this study, model polyals were produced via lateral cleavage of polyaldoses and polyketoses. Best results were achieved using dextran B-512 as a precursor. The resultant poly[hydroxymethylethylene hydroxymethylformal], in agreement with the hypothesis, demonstrated excellent biological properties and technological flexibility. Materials of this type can potentially have several applications in pharmacology and bioengineering.