In the course of a study of UDP-xylose:proteoglycan core protein xylosyltransferase (EC 2.4.2.26), another xylosyltransferase was discovered in the soluble fraction of a rat kidney homogenate. The latter enzyme catalyzed [3H]xylosyl transfer from UDP-[3H]xylose to an endogenous acceptor and yielded a product in which the xylose was bound by an alkali-stable linkage. It was therefore concluded that the acceptor was not the core protein of one of the proteoglycans containing a xylose-->serine linkage, since this linkage is cleaved by alkali. The [3H]xylose-labeled product emerged with the void volume when chromatographed on Sephadex G-50, it was precipitated by trichloroacetic acid, and it had a mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis corresponding to a molecular mass of about 32,000 Da. Digestion with trypsin or alpha-amylase degraded the labeled product to small fragments, as determined by gel chromatography, suggesting that it was a glycoprotein related to glycogen. A product of similar characteristics was formed when UDP-[3H]glucose was substituted for UDP-[3H]xylose as the glycosyl donor, and the two nucleotide sugars were mutually competitive in the respective transfer reactions, indicating that they were substrates for the same enzyme. On the basis of these findings, it was tentatively concluded that the xylosyltransferase and its acceptor were the renal form of glycogenin.