BackgroundThe British horseracing industry is committed to reducing equine fatalities in jump racing. Race‐related fatalities are a major welfare concern and threaten the sport's social licence to operate.ObjectivesTo describe the risk of, and determine risk factors for, fatality in British jump racing.Study designRetrospective cohort.MethodsAnalyses included all starts made in British jump races between January 2010 and April 2023. Available information for each horse, race, racecourse, trainer and jockey was collated and combined with details of all fatalities recorded by official veterinary officers in a central database. A fatality was defined as any post‐start veterinary event that resulted in the sudden death or euthanasia of a horse within 48 h of racing. Risk factors (n = 101) were evaluated using mixed‐effects logistic regression. Data for steeplechase and hurdle starts were analysed separately.ResultsThe overall fatality rate was 5.9 per 1000 steeplechase starts (n = 836/141 922; 95% CI 5.5–6.3) and 4.5 per 1000 hurdle starts (n = 1096/242 486; 95% CI 4.3–4.8). In both race types, fallers (steeplechase: OR 28.7, 95% CI 23.0–35.8; hurdle: OR 41.4, 95% CI 32.9–52.0) and older horses (steeplechase: OR 1.1 per extra year, 95% CI 1.1–1.2; hurdle: OR 1.2 per extra year, 95% CI 1.1–1.2) had higher odds of fatality. In steeplechase racing, starts made in summer (OR 1.2, 95% CI 1.0–1.5) and by non‐GB trained horses (OR 2.0, 95% CI 1.4–3.0) experienced higher fatality odds. In hurdling, maiden races (OR 1.3, 95% CI 1.0–1.6) were at higher odds of fatality. In both race types, softer going decreased the odds of fatality. Approximately half of the unexplained variation in fatality odds was attributable to horse.Main limitationsAnalysis was restricted to routinely recorded race‐day factors and performance history.ConclusionsReducing the risk of falling and racing on softer ground could substantially decrease fatalities in jump racing.