The objectives of the study were to develop information regarding the frequency of recurrence of exercising arrhythmias and the relationship of arrhythmia development to exercise intensity and type of exercise in Thoroughbred horses. Electrocardiograms (ECGs) were recorded on nine Thoroughbreds during maximal or submaximal exercise on a racetrack (Ra) and treadmill (Tm). The frequency of arrhythmias on a Ra and Tm was compared, and their relationship to exercise intensity (expressed as HR/HRmax [%]) was evaluated. Sixty-five workouts were analyzed: 46 workouts were on a Tm and 19 on a Ra; median number of workouts/horse was four, and the range was 2–14. Exercising arrhythmias were detected in 4/9 horses (12/65 workouts), and there were postexercise arrhythmias in 7/9 horses (19/65 workouts). Arrhythmias were detected at some point in 8/9 horses. For 7/9 horses, the same rhythm result was obtained during exercise in repeated recordings. For 7/9 horses, the postexercise rhythm was variable: postexercise arrhythmias were present in median: 21%; range: 0%–75% of workouts. The presence of arrhythmias was positively related to exercise intensity (P = .01; odds ratio = 1.2) and all occurred during workouts at ≥94% of HR/HRmax (%). Arrhythmias during exercise were more frequent on the Ra than on the Tm (P = .009). A single ECG did not always display all the arrhythmias detected over several exercise tests. The presence/absence of exercising arrhythmias was more consistent than postexercise arrhythmias. Arrhythmias were more likely to be detected at maximal or near-maximal intensities and during gallops on the Ra. A larger population needs to be studied before more definitive conclusions are drawn.