Cardiac arrhythmias resulting in sudden cardiac death (SCD) are an important cause of racehorse fatalities. The objective of this study was to determine risk factors for SCD in Thoroughbreds by evaluating a sample with a policy of mandatory post-mortem following racing or training fatalities. Risk factors were compared between case horses with SCD ( = 57) and control horses with other fatal injury (OFI, = 188) by univariable and multivariable logistic regression. Survival in years for horses with SCD was compared to OFI using the Kaplan-Meier method with log rank test. The following variables were most important in the multiple logistic model: Horses with SCD were more likely to die during training than during racing, SCD (42/57, 74%) vs. OFI (82/188, 44%; odds ratio [OR], 95% confidence interval [CI], 2.5, 1.2-5.4; = 0.01), had fewer lifetime starts, median (interquartile range [IQR]), SCD (3.0 [0.0-9.0]) vs. OFI (9.0 [0.0-22.8]; OR, 95% CI, 0.96, 0.9-1.0; = 0.02 and were less likely to be entire (uncastrated) males, SCD 9/57 (16%) vs. OFI (46/188, 25%; OR, 95% CI, 0.47, 0.1-0.9; = 0.03). Survival in years (median (IQR)) for horses with SCD was 3.6 (3.1-4.4), which was shorter than OFI (4.5 [3.1-6.0], hazard ratio, 95%CI, 1.6,1.2-2.3; < 0.001). SCD occurs more commonly in training than racing, which suggests exercise intensity is less important in precipitating this fatality. In this study, SCD occurred early in the careers of affected horses.n