This study tested our hypothesis that during an 80-km Research Ride in 2002 (R2) horses that did not finish (NF) the ride would have elevated muscle enzyme activities in the blood and changes in biomarkers of oxidative stress as compared to horses that finished (F) the ride. These results were then compared to previous rides – Old Dominion (OD) and the Research Ride 2001 (R1). For R2, 40 mostly Arabian horses competed and had blood samples collected before, at 27, 48 and 80 km, and 170 to 190 min after the 80-km race. Blood was collected similarly in R1 and OD. Blood was analysed for plasma lipid hydroperoxides (LPO), α-tocopherol (TOC), creatine kinase (CK), aspartate aminotransferase (AST), red and white blood cell total glutathione (GSH-T) and glutathione peroxidase (GPx). Data were analysed using a repeated measure ANOVA in SAS. Associations between muscle enzymes and antioxidant status were determined using Pearson's or Spearman's correlations. Activities of CK and AST were higher (P<0.05) before, during and after the ride in NF than in F; however, TOC, LPO, GSH-T and GPx were not different. In R2, negative correlations were found with GPx and CK (r = −0.21; P = 0.005), GPx and AST (r = −0.15; P = 0.05), and a positive correlation was found with GSH-T and CK (r = 0.18; P = 0.02). Values of CK, LPO, GPx and GSH-T were higher (P<0.05) in R2 than in R1 or OD. The overall comparison of 80-km endurance races suggests the importance of considering the horse's fitness, terrain, ambient conditions and calibre of race when interpreting results from markers of oxidative stress and muscle enzyme leakage.