BACKGROUND: Heart rate variability (HRV) analysis measures the inter-beat interval variation of successive cardiac cycles. Measurement of these indices has been used to assess cardiac autonomic modulation and for arrhythmia identification in exercising horses. OBJECTIVES: To report HRV indices during submaximal exercise, strenuous exercise and recovery, and explore relationships with clinical conditions (arrhythmias, lameness, equine gastric ulcer syndrome [EGUS], lower airway inflammation and upper respiratory tract obstructions [URTOs]) in Thoroughbred racehorses. STUDY DESIGN: Retrospective, observational cross-sectional study. METHODS: One hundred and eighty Thoroughbred horses underwent a treadmill exercise test with simultaneous electrocardiographic recording. Time-domain HRV indices (standard deviation of the R-R interval [SDRR]; root mean square of successive differences [RMSSD]) were derived for submaximal and strenuous exercise and recovery segments. Clinical conditions (arrhythmia [during each phase of exercise], lameness, EGUS, lower airway inflammation and URTO) were assigned to binary categories for statistical analysis. Relationships between selected HRV indices and the clinical conditions were explored using linear regression models. RESULTS: During submaximal exercise, lameness was associated with decreased logRMSSD (B = -0.19 95% confidence interval [CI] -0.31 to -0.06, p = 0.006) and arrhythmia was associated with increased logRMSSD (B = 0.31 95% CI 0.01-0.608, p = 0.04). During strenuous exercise, arrhythmia was associated with increased HRV indices (logSDRR B = 0.51 95% CI 0.40-0.62, p < 0.001; RMSSD B = 0.60 95% CI 0.49-0.72, p < 0.001). During recovery, arrhythmia was associated with increased HRV indices (logSDRR B = 0.51 95% CI 0.40-0.62, p < 0.001, logRMSSD B = 0.60 95% CI 0.49-0.72, p < 0.001). MAIN LIMITATIONS: The main limitations of this retrospective study were that not every horse had the full range of clinical testing, therefore some horses may have had undetected abnormalities. CONCLUSIONS: The presence of arrhythmia increased HRV in both phases of exercise and recovery. Lameness decreased HRV during submaximal exercise.