Arthritis refers to a heterogeneous class of diseases characterized by impairment of movement. Yet animal models of arthritis have traditionally been based on the utilization of animals housed without the capability of extended free movement and without adjunctive measurement of mobility. To define the determinants of mobility impairment, we have established a lipopolysaccharide (LPS)-induced arthritis model in the hamster that prominently features monitoring of mobility and compares mobility changes with histological and biochemical changes during arthritis. Intraarticular LPS induces a dose-dependent inhibition of the hamster's mobility as measured by decreased daily distance on a running wheel (normal distance 9 to 12 km/day). At low concentrations of LPS (0.1 and 1 microgram/knee), daily distances returned to normal after 4 and 6 days, respectively. At higher concentrations, the mobility was still markedly suppressed after 6 days, and, at 100 micrograms/knee, irreversible chondrocyte loss was observed on the femoral condylar margins. Further studies were therefore conducted using 1 microgram LPS/knee. Histological and biochemical changes were examined to determine which resolved at the time of restoration of mobility. At the time of restoration of mobility, the synovial capsule was still edematous and heavily infiltrated with leukocytes; proteoglycan loss from the medial femoral condyle was still increasing. Plasma keratan sulfate failed to correlate with either proteoglycan loss or mobility changes. Proteoglycan synthesis, which was maximally suppressed the second day after LPS, was enhanced over controls at the time of restoration of mobility, suggesting the onset of repair. These results suggest a possible association of mobility inhibition with local cytokine synthesis. This model provides an approach to define the causes of mobility impairment.