A randomised double blind and placebo controlled clinical study was carried out. Standardbred trotters (n=77), age 3-4 years) with moderate to severe lameness were grouped according to number of affected joints and, within each group, were randomised for treatment with polysulphated glycosaminoglycan (PSGAG), sodium hyaluronate (SH) or placebo for 3 weeks. The horses were inspected weekly with a final examination 2-4 weeks after the end of treatment. Mean initial lameness score was significantly reduced during treatment and at the last examination in all 3 groups (P<0.01). Additionally, the prevalence of sound horses increased significantly from 1 to 3 weeks of treatment and to the last examination in all 3 groups (P < or = 0.03). Comparison of the 2 treatment groups with regard to development of the lameness curve and time until soundness indicated a small, non significant difference in favour of SH. No significant difference was detected between the 2 treatment groups in the prevalences or cumulative incidence of soundness. The study detected a superior effect of the 2 drugs compared with placebo for reduction in lameness score during the treatment period (P=0.03) and the total study period (P<0.01), time until soundness (P=0.04) and the prevalence of sound horses at the last examination (P<0.01). All 3 treatments affected traumatic arthritis in horses, but the SH and PSGAG gave better results than the placebo.