Information on equine stress levels are important in evaluating many aspects of horse management and training. For example, for correct assessment of equine welfare, implications of training and handling procedures, success of training and the interpretation of equine temperament. Stress is a generic concept describing the body's reaction to external stimuli, including both physiological and psychological factors. Therefore, by definition, the assessment of psychological stress in the exercising horse encompasses the problem of teasing apart the psychological and physiological factors both of which result in stress responses.. The present paper reviews the existing literature on various measures of stress taken specifically in the context of equitation science. Particular attention has been paid to short-term effects, and commonly used measurements of short-term stress include heart rate, a number of heart rate variability parameters, blood or saliva cortisol levels, eye temperature, and various behaviour parameters including in particular behaviour patterns presumably indicative of conflict with the rider's/trainer's aids. Inspection of the individual studies' results revealed that disagreement between these different measures of stress is commonplace. For physiological parameters, the largest proportion of agreement (i.e. both parameters simultaneously indicated either higher, non-significant or lower stress compared to a control treatment) was found for heart rate and heart rate variability parameters, while generally limited agreement was found for cortisol and cardiovascular parameters. It appears that cortisol levels may not be particularly useful for assessment of the valence of a situation in the exercising horse as cortisol levels are predominantly linked to activity and exercise levels. Although heart rate variability parameters reflect in theory more closely sympathovagal balance compared to cortisol levels, great care has to be taken regarding the use of appropriate equipment, time-frames, as well as raw data correction. In spite of its widespread and apparently successful use, popular equipment may in fact not be accurate enough under 2 field conditions. Eye temperature is another promising parameter for assessment of psychological stress, but the technique is likewise susceptible to application errors. Given the high susceptibility of physiological parameters to errors at various experimental stages, behavioural rather than physiological parameters may in fact provide more accurate measures of valence when conducting experiments in the exercising horse. Behavioural parameters that appear to be particularly practical in assessing stress in ridden horses' behaviour are frequencies of behaviour patterns indicative of conflict. However, while increased frequencies of behaviour patterns indicative of conflict are a good indicator of stress, the absence of conflict behaviour does not provide proof of the absence of stress due to the possible occurrence of conditions such as Learned Helplessness. In future...