CH64 7TEFALLS ofhorse and rider are rare events, having an incidence of one fall per 1160 jumping efforts (Singer and others 2003). To date, the prevalence and nature of injuries sustained by event horses as a result of falling has not been documented fully. Common injuries to horses resulting from falls have been reported to include lacerations, neck trauma, and fractures of the supraglenoid tubercle and pelvis (Dyson 1996). Unpublished data from British Eventing (BE) records show an average incidence rate of 2-3 horse fatalities per 10,000 crosscountry rounds for the years 1996 to 1999. The aims of this study were to provide an estimate of the relative risk of injury to the horse associated with falls at different types of fence.Horse falls that occurred during the 2000 eventing season were identified from BE cross-country score sheets. A horse fall is defined as an incident in which the shoulders and hindquarters of the horse touch the ground or the obstacle and the ground at the same time (British Horse Trials Association 2000). Information on variables relating to horse falls was obtained from the event programmes. The variables were event level (prenovice, novice, intermediate or advanced), competition type (one, two or three-day event), obstacle design (for example, upright or ascending spread) and obstacle construction, for example, filled (a solid wall) or open (a post or rail fence). A telephone questionnaire was undertaken with all riders of horses that had fallen, to verify the BE information and to document additional details of the fall, including injuries to the horse and the time it took for the animal to recover from the fall.In calculating the risk of injury associated with a fall, both injury (all types) and specific injuries were considered as outcomes. The relationship between injury and risk factors was examined using Mantel-Haenszel chi-squared summary statistics. The level of significance was set at P<0*05 for all univariable analyses.A total of 424 falls of horse and rider were identified from the BE score sheets. Of the 424 falls identified, the riders confirmed 345 (81.4 per cent) of these as horse falls. Thirty-nine riders (9-2 per cent) were unable to be contacted to confirm the fall, and 40 of the recorded falls (9.4 per cent) were misclassified, according to the rider. All 345 riders who confirmed a fall of horse and rider completed telephone questionnaires. 80 70 o, 10 -$ 8 6 4 2 o0 None <11-2 3-4 5-8 9-12 >12 Horse euthanased From the sample of 345 riders, 228 (66-1 per cent) reported that their horse had sustained no injuries, 112 (32-5 per cent) reported one or more injuries and information regarding injury was unobtainable for five (1-5 per cent) of the riders. A summary ofthe types and locations of injuries recorded is given in Table 1. The prevalence of injury among fallers was not significantly related to the level of the event for horses competing at one-day events. No similar analysis was performed for the risk of injury at different levels of three-day events, as the numb...