The structure of rat liver heparan sulfate (HS) has been investigated using a combination of (a) chain scission with specific reagents, (b) disaccharide compositional analysis, and (c) end-referenced sequence analysis of the proximal, protein-linked region of the chain. This study reveals that the liver synthesizes a highly sulfated HS species (1.34 sulfates/disaccharide), particularly high in N-sulfation (60%) and 2-O-sulfate content (36%). Approximately half of the latter is found in trisulfated disaccharides, i.e. IdceA(2-OSO3) alpha 1-4GlcNSO3 (6-OSO3). End-referencing methodology established the existence of an extended, unmodified heparan (GlcUA beta 1-4GlcNAc) sequence, 8-11 disaccharides in length, attached to the linkage tetrasaccharide, similar to that found in a number of other HS species. Directly following this is a mixed HexUA1-4GlcNR(6-OSO3) (where GlcNR represents alpha-D-glucosamine with an unspecified N-substituent)-containing sequence of variable length, culminating in the appearance of the first IdceA(2-OSO3) residue approximately 20 disaccharides from the linkage region, i.e. approximately 40% along the length of the chain. The distal 60% of the polysaccharide is highly sulfated (approximately 2 sulfates/disaccharide) and mainly comprises three heparin-like domains, highly enriched in IdceA(2-OSO3) residues. Overall, liver HS qualifies as an extreme member of the HS family, with a considerable proportion of heparin-like structure asymmetrically concentrated to the distal part of the chain.