An investigation was made into the possible contribution of autoimmune mechanisms to equine arthropathies. Serum and synovial fluid (SF) immune complexes and anti-collagen Type II antibodies were measured, by ELISA, in groups of horses with naturally occurring osteoarthritis (OA), osteochondritis dissecans (OCD), bone fracture, traumatised joints, synovitis, infected joints and non-diseased (control) joints. Significantly raised anti-collagen Type II antibodies were found in osteoarthritic (P < 0.02) and traumatised joint synovial fluids (P < 0.01) compared with the control, where ten of 38 (26%) OA and five of eight (63%) traumatised joint synovial fluid samples had raised anti-collagen Type II antibodies (above mean and 2 SD of control group; P < 0.05). Significantly raised levels of C1q-binding immune complexes were found in the synovial fluids of horses with OA (P < 0.001), OCD (P < 0.02), fractured articular bone (P < 0.001), infected (P < 0.01) and traumatised joints (P < 0.001) as compared with the control horses. Raised C1q-binding activity (above mean and 2 SD of control group; P < 0.05) was shown in synovial fluids in 18 of 38 (47%) OA, four of 16 (25%) OCD, four of nine (44%) fractured bone, one of eight (14%) synovitis, two of four (50%) infected joints and seven of eight (88%) traumatised joints. No raised anti-collagen Type II antibodies or C1q-binding activity were detected in sera from any clinical group. Correlations were found, in synovial fluids, between anti-collagen Type II antibodies and C1q-binding activity in OA (P < 0.02), synovitis (P < 0.01) and infected joint (P < 0.05) groups. Within individual horses, there were no correlations between the SF and sera measurements of either anti-collagen II or C1q-binding activity. The relationship between cause and effect of these immunological findings cannot be determined; as they are common to many types of equine joint disorder it is probable that they are not an initiating factor in pathology.