The agarose-polyacrylamide gel electrophoresis procedure for the analysis of proteoglycans originally described by C. A. McDevitt and H. Muir (1971, Anal. Biochem. 44, 612-622) has been modified to minimize trailing and to allow the analysis of crude samples, i.e., tissue extracts. A slab gel system was used, permitting reproducible analysis of many samples. Procedures are described that can be used to separate and quantify several subpopulations of proteoglycans and also to quantify the proportion of proteoglycans capable of aggregating with hyaluronic acid. Applications of the procedure include transfer to nitrocellulose paper followed by immunological detection of proteoglycans as well as fluorography of separated, radiolabeled proteoglycans.