The hydrolytic activities of beta-galactosidases from three different sources have been determined in various 50% (v/v) organic solvent-buffer mixtures with a view to finding solvent systems of reduced water content suitable for the synthesis of glycosides and oligosaccharides. K. fragilis beta-galactosidase lost activity rapidly (within 30 min) in all mixtures examined. The enzymes from E. coli and A. oryzae retained significant activity during extended periods (120 h) and are suitable for synthetic experiments. The level of activity depended on the enzyme and organic component of the solvent in the order tetraglyme > triethyl phosphate > monoglyme > triglyme = trimethyl phosphate > diglyme > diethylene glycol diethyl ether > 2-methoxyethyl acetate > acetone = acetonitrile (zero activity) for the E. coli enzyme and methoxyethyl acetate > trimethyl phosphate > triethyl phosphate > monoglyme > acetone approximate to diglyme approximate to triglyme approximate to tetraglyme > diethyleneglycol diethyl ether > acetonitrile for the A. oryzae enzyme. The decay in hydrolytic activity over time was exponential, and in some cases biphasic with the second decay phase being an order of magnitude slower than the first. The reduction in enzyme activity was shown to be reversible with respect to the removal of organic solvent by dialysis and is attributed to changes in substrate solvation as well as partial enzyme denaturation. Changes in V-max/K-M for the beta-galactosidase-catalysed hydrolysis of 2-nitrophenyi-beta-D-galactopyranoside (ONPGal) are weakly correlated with lipophilicity of the added cosolvent. In several cases, we successfully used various organic water-miscible solvent systems in the aid of oligosaccharides versus hydrolysis. (C) 2004 Elsevier Inc. All rights reserved.