Keratan sulfate (KS) is a glycosaminoglycan, distribution of which is confined mostly to hyaline cartilage. As such, it is a putative marker of hyaline cartilage catabolism. In experiment 1, a focal osteochondral defect was made arthroscopically in 1 radial carpal bone of 2 ponies, and in 2 other ponies, chymopapain was injected into the radiocarpal joint to induce cartilage catabolism. Sequential and concurrent plasma and synovial fluid concentrations of KS were measured, up to 13 months after induction of cartilage injury, to determine whether changes in KS concentrations reflected cartilage catabolism. In experiment 2, a large, bilateral osteochondral defect was made in the radial carpal bones of 18 ponies, which were subsequently given postoperative exercise and/or injected intra-articularly with 250 mg of polysulfated glycosaminoglycan (PSGAG). Medication was given at surgery, then weekly for 4 weeks. Blood samples were collected and synovial fluid was aspirated before surgery, when medication was given, and at postmortem examination (postoperative week 17). The KS concentration was measured in these fluids to determine whether changes in KS concentration indicated an effect of joint treatment. In experiment 1, the concentration of KS in synovial fluid was highest 1 day after joint injury, and the concentration in plasma peaked 2 days after joint injury. For ponies receiving chymopapain intra-articularly (generalized cartilage catabolism), a fivefold increase over baseline was observed in the concentration of KS in plasma (peak mean, 1.2 microgram/ml), and a tenfold increase over baseline in synovial fluid (peak mean, 2.0 mg/ml) was observed. On average, these maxima were threefold higher than values in fluids of ponies with osteochondral defects (focal cartilage disease). In experiment 2, nonexercised ponies had lower KS concentration (as a percentage of the preoperative concentration) in synovial fluid than did exercised ponies at all postoperative times, and at postoperative week 17, this effect was significant (P < 0.05). This may be related to decreased turnover of KS in articular cartilage attributable to stall confinement and late increase in turnover related to exercise. Seventeen weeks after surgery, synovial fluid from exercised, medicated ponies had significantly (P < 0.05) higher KS content than did fluid from exercised, nonmedicated ponies. This indicated that exercise, when combined with medication, may increase KS release from articular cartilage. Synovial fluid from medicated joints of nonexercised ponies had significantly (P < 0.05) lower KS concentration than did synovial fluid from nonmedicated joints of nonexercised ponies. This indicated that, in nonexercised joints, medication with PSGAG may have decreased either release of KS from the articular cartilage into the synovial fluid or inhibited synthesis of KS. Concentration of KS in synovial fluid was not related clearly to the development of osteoarthritis in these ponies. Exercise or medication did not affect plasma KS concentration, and synovial fluid and plasma KS concentrations were not correlated. Data indicated that KS concentration in plasma and synovial fluid may be increased in acute, marked, generalized articular cartilage catabolism and that KS turnover in cartilage of joints with large osteochondral defects was affected by intra-articular PSGAG and postoperative exercise.