Eventing safety: the fall data, and how frangible fences came in.

The deaths and falls on this page are recorded as published history and statistics, not as lessons. Any injury from a fall needs medical care. For how eventing is scored and what the cross-country penalties are, see the eventing page.

The FEI's international statistics, 2016 to 2025

The FEI's risk management deck "2025 Highlights of international statistics" was presented at the FEI Online Eventing Seminar on 31 January 2026 by David O'Connor, chair of the FEI Eventing Committee, and Stuart Buntine, chair of the Risk Management Steering Group. We read the copy US Equestrian hosts. It covers international eventing only.

• Falls rise with level. In 2025 falls were about 4.2 to 4.3 percent of starters at one and two star, 5.09 percent at three star, 6.43 percent at four star and 9.90 percent at five star. Long-format competitions averaged 6.30 percent over 2016 to 2025, against 4.88 percent for short format.
• Unseated riders and horse falls. Over the ten years the FEI averaged 4.00 unseated athletes per 100 starters (one in 25) and 1.22 horse falls per 100 starters (one in 82). In 2025 the figures were 3.96 and 0.92 percent.
• Rotational falls. Carol Hall (The Conversation, 2016) describes these as falls in which the horse hits a fence and somersaults. They fell from 30 in 2016 (one in 664 starters) to 15 in 2025 (one in 1,336), and the FEI says they have decreased steadily since 2016.
• Where falls happen. In 2025, 72.19 percent of falls were at cross-country fences and 27.81 percent elsewhere, in dressage, jumping, warm-up and other places.
• Injuries. In 2025, 86.81 percent of falls caused no injury, 10.94 percent slight injury and 2.25 percent serious or fatal injury. The FEI reported 22 serious or fatal injuries and 22 concussions. Of the 15 rotational falls, 3 led to a serious injury.
• Which falls hurt most. Averaged over the period, an unseated athlete suffered a serious injury about once in every 76 falls, against about once in every 5 rotational horse falls.

Since 2019 the FEI has defined a serious injury as one that, in the referring doctor's opinion, needs hospital admission for immediate treatment. The FEI also sets target horse-fall rates by level and format, from 0.50 percent at CCI2*-S up to 4.00 percent at CCI5*-L; the 2025 CCI5*-L figure was 3.30 percent.

The deck combines serious and fatal injuries in one category, so it gives no separate count of deaths. Older FEI fatality rates quoted in Australian coverage could not be confirmed and are not given here.

The peer-reviewed studies

Bennet, Cameron-Whytock and Parkin (2021). Euan Bennet, Heather Cameron-Whytock and Tim Parkin, of Bristol Veterinary School and Nottingham Trent University, published an FEI-funded study in the Equine Veterinary Journal covering all 202,771 horse starts at FEI international, championship, Olympic and World Equestrian Games eventing from 2008 to 2018; 187,602 of them started cross country. A horse fell in 1.5 percent of cross-country starts and a rider was unseated in 3.5 percent. Higher odds of a horse fall were linked with a higher level, a longer course, more starters, mares compared with geldings, and fewer starts at the current level. Athlete factors included being male, younger, less experienced and not finishing the previous event. Pairs with a dressage penalty score above 50 had higher odds of a cross-country fall.

O'Brien and Cripps (2008). Denzil O'Brien and Raymond Cripps, of the Flinders University Research Centre for Injury Studies, reported to Australia's Rural Industries Research and Development Corporation on 1,732 falls at 444 Australian events from 2002 to 2006, and on 25 rider deaths worldwide from May 1997 to September 2007. In their data, 18 of the 25 deaths followed rotational horse falls, and 17 of those riders were crushed by the horse. They named reducing horse falls as the most effective way to reduce injuries. Of 611 surveyed riders who called their own fall preventable, all but 60 put it down to their own riding, and 18 blamed the horse.

How the safety response unfolded

1999 and the Hartington Report

In the 1999 season five event riders died in Great Britain. Carol Hall of Nottingham Trent University, writing in The Conversation in 2016, adds one death in the US that year and says five of the six were rotational falls; Practical Horseman's 2018 history gives the five British deaths only, so the two counts are compatible. The deaths led to the FEI International Eventing Safety Committee, chaired by the Marquess of Hartington. Its report of April 2000 named frangible technology and worldwide fall data as the two most critical areas.

The sources disagree on when the FEI's fall data began. Practical Horseman says the FEI began a worldwide falls database in 2002, says elsewhere in the same article that FEI data recording began in 2006, and dates the first FEI annual eventing statistics report to 2007.

Frangible pins in Britain

A frangible device lets part of a fence give way under enough force, so the fence collapses instead of staying rigid. Practical Horseman reports that British Eventing and the Transport Research Laboratory trialled frangible pins on post-and-rail fences at nine horse trials, including Badminton, in 2002. Another account says pins were available from 2001; that version is unconfirmed. Hall writes that from 2006 British Eventing required pins on as many fences as possible, while at FEI level their use was then left to the course designer.

2007 and 2008

Animal People reported in June 2008, citing the Lexington Herald-Leader, that at least 12 riders had died worldwide in about 18 months, all on cross country. That month US Equestrian and the US Eventing Association met in Lexington, and USEF president David O'Connor said safety could be improved by reducing horse falls. Practical Horseman records that at the 2008 Red Hills Horse Trials in Florida two horses died of pulmonary haemorrhage and rider Darren Chiacchia was seriously injured in a rotational fall, after which the USEA formed a Cardiopulmonary Research Group led by Dr Catherine Kohn.

Clips, ratings and the 2021 rule

• 2013: the FEI approved the MIM Safe New Era Clip, which releases under both vertical and horizontal force. MIM clips were developed by Swedish inventor Mats Bjornetun, whose company also tests car safety devices.
• 2015 to 2018: Practical Horseman reports that EquiRatings launched in 2015, that Ireland became the first national federation to use its quality index in 2016, that the FEI signed a four-year agreement with EquiRatings in 2017, and that the USEA launched the index nationally in 2018. It also reports rotational falls falling from about one in 268 starters to about one in 572 after FEI data recording began.
• 6 April 2020: version 2.5 of the FEI frangible device standard. Devices made to the older Version 1 standard could be used until 31 December 2021.
• 1 January 2021: under the FEI Eventing Rules approved by the FEI General Assembly in November 2020, open oxers, open corners, verticals with open rails, triple-bar top rails and gates at CCI1* to CCI5* must carry FEI-approved frangible devices where the rails fit the device parameters.

In the 2026 FEI Eventing Rules, as reproduced in Equestrian Australia's 2026 rules (Art 548), activating a frangible device that changes the fence's dimension costs 11 penalties. The figure was 21 until 2015: Eventing Nation reported on 30 March 2015 that the FEI had revised the rule, from 31 March 2015, so that activating a frangible device cost 11 penalties instead of 21. A search summary of Horse and Hound says British Eventing gives no penalty for frangible activation in domestic classes; we could not read the current British Eventing rulebook, so that is unconfirmed. Reports that Equestrian Australia and US Equestrian made frangible devices mandatory on specified fences in 2018 are also unconfirmed.

Do frangible devices prevent falls?

The sources do not say the same thing.

• Carol Hall (The Conversation, 2016) cites an FEI audit finding that fences with frangible pins were associated with a higher risk of falls, meaning collapsible fences alone do not prevent falls. Advocates, she notes, describe the devices as reducing severity rather than preventing falls.
• The Chronicle of the Horse (March 2026) reports that frangible devices are associated with a statistically significant reduction in rotational falls, and that a streamlined pathway for approving frangible designs at national level was announced in 2025.

The two are not measuring the same thing: one is about falls in general, the other about rotational falls. For how bodies handle concussion after a fall in competition, see concussion protocols, and for injury data beyond eventing, see what the injury data shows.