We investigated agreement between eye maximum pixel temperature (using thermography) and rectal temperature (TR) in racehorses, comparing the results with salivary cortisol concentration and heart rate (HR), both at rest and after exercise. Nineteen horses, undergoing training for racing in their first racing season, were studied. Eye maximum pixel temperature, TR, salivary cortisol concentration, and HR were measured before training (BT), within 5 minutes of the end of the training session (T+5), and 2 hours after training (T+120). Eye maximum pixel temperature, TR, salivary cortisol concentration, and HR were all significantly elevated at T+5 compared to BT (all P < .001). At T+120, only eye maximum pixel temperature remained significantly elevated compared to BT (P < .05). Bland-Altman analysis indicated a poor agreement between eye maximum pixel temperature and TR. We noted no significant correlations among any of the measurements at any time point, with the exception of eye maximum pixel temperature and TR at BT (r = 0.55, P = .01). In racehorses, eye maximum pixel temperature is a poor estimate of core temperature due to limited agreement with TR. Furthermore, eye maximum pixel temperature is not correlated with accepted measures of stress such as salivary cortisol concentration and HR.