Artificial substrates such as 4-methylumbelliferyl glycosides and p-nitrophenyl glycosides have been employed commonly to determine the levels of glycosyl hydrolase activities in plasma, urine and tissues [l-7]. They have provided a simple and rapid method for the diagnosis of some lysosomal storage diseases, whereas in other instances inconsistencies have been noted in the results obtained with these artificial substrates and with the natural substrates. Two examples which may be cited are Krabbe’s globoid cell leucodystrophy, characterized by a galactocerebroside fl-galactosyl hydrolase deficiency, for which p-nitrophenyl P_galactoside cannot be used to detect decreased enzymatic activity in the homozygous patients [8] ; and the isolation from Tay-Sachs brain of hexosaminidase which has a much higher Km for the GM2 ganglioside substrate than for the p-nitrophenyl P-N-acetylgalactosaminide used to monitor the enzyme during purification [9]. We have reported preliminary studies of a-galactosidase activity in Fabry plasma following infusion of normal plasma, and suggested that the enzymatic activities measured by the artificial substrates and by cr-galactosyl-( 1 --f 4)flgalactosyl( 1 --f 4)glucosylceramide (ceramide trihexoside) are not due to the same enzyme. The present report provides unequt vocal experimental evidence for the existence of unique cu-galactosidases for the artificial substrates and for ceramide trihexoside. 2. Experimental