The incidence and severity of exercise-induced pulmonary haemorrhage (EIPH) in the 2 most commonly raced horse breeds, Thoroughbreds (TB) and Standardbreds (STD), were studied, with particular interest in the possible influence of frusemide (F) and/or the breed (or running gait) on EIPH. The appearance of blood within the trachea was semi-quantified using a published 5-point system, with zero assigned when no blood was observed, and numbers 1-4 assigned with increasing amounts of blood. Considering each endoscopic examination as a separate event, approximately 75% of the postrace endoscopic examinations had blood-scores of 1, 2, 3, or 4, regardless of breed or F administration. For horses examined twice, the chances of finding blood-scores of 1 or greater in either of the examinations increased to approximately 95%. All horses examined 3 or more times had endoscopic blood-scores of 1 or greater following one or more races, again, irrespective of the breed or F administration. Mean +/- s.e. 'blood scores' were 1.5 +/- 0.1 and 1.8 +/- 0.2 for TB, and 1.4 +/- 0.2 and 1.2 +/- 0.1 for STD racing with and without prerace F, respectively. Therefore, there was no apparent effect of breed (or possibly racing gait) on EIPH, and no differences in the incidence or severity of EIPH were observed between horses with or without prerace frusemide administration.