Studies of the initiation of the polysaccharide chains in proteoglycans containing heparins, heparan sulfates, or chondroitin sulfates — all linked to the core proteins of the respective proteoglycans by a xylose→serine linkage — began about 30 years ago. As had been intended, these investigations led to the demonstration of transfer of xylose from UDP-xylose to serine hydroxyl groups in an acceptor of protein nature, which was presumably the core protein of the proteoglycan under study. This reaction is catalyzed by the enzyme proteoglycan core protein xylosyltransferase. However, they also resulted in the discovery, in extracts of a mouse mastocytoma and hen oviduct, of a second xylosyltransferase which gave a product of a different kind. The property that set this product apart was its resistance to alkali treatment under conditions that cleave the xylose→serine linkage in the proteoglycans. We have found an analogous xylosyltransferase in kidney and muscle, and extensive characterization of the alkali-stable product formed by a soluble enzyme preparation from rat kidney has led to the conclusion that the enzyme (and the acceptor!) are identical to glycogenin. This substance is the core protein of glycogen proteoglycan and is also a self-glycosylating enzyme that normally uses UDP-glucose as the glycosyl donor and catalyzes early steps in glycogen biosynthesis. In this review, we summarize our current knowledge of the polysaccharide chain initiating enzymes glycogenin and proteoglycan core protein xylosyltransferase and how the discovery that both enzymes can use the common substrate UDP-xylose has led to new insights into glycogen and proteoglycan biosynthesis and its regulation. © 1995, FCCA(Forum: Carbohydrates Coming of Age). All rights reserved.