Four groups of 8 horses each had 1 midcarpal joint injected with 33 colony-forming units (CFU) of viable Staphylococcus aureus plus: 1 ml of saline solution (group 1, control), 250 mg of polysulfated glycosaminoglycan (PSGAG, group 2), 100 mg of methylprednisolone acetate (group 3), or 20 mg of sodium hyaulronate (group 4). Horses were euthanatized, and samples were obtained on the basis of clinical signs of septic arthritis that were nonresponsive to phenylbutazone administration. One group-1 horse, all 8 group-2 horses, 3 group-3 horses, and 4 group-4 horses were culture-positive for S aureus and had clinical signs, results of synovial fluid analysis, and histopathologic findings that were consistent with sepsis. The addition of 250 mg of PSGAG increased the development of sepsis significantly (P = 0.001), compared with results in control horses. Differences in the development of sepsis between horses injected with methylprednisolone acetate or sodium hyaluronate and control horses were not significant.