Abstract: The biological diversity of a species gives rise to individual differences in behavioural tendency. Traditionally, this variation has been considered to be of little scientific importance or value, but the description and quantification of the fundamental basis of this variability is relevant to many aspects of equine science. The reliable identification of these features may allow the development of more accurate diagnostic and prognostic indicators for a range of clinical diseases. Biologically based traits also provide a more rational basis for selective management and breeding programmes in which specific behavioural tendencies are sought. Individual differences in behaviour also reflect the range of subjective feelings experienced by animals and therefore need to be understood by those concerned with animal welfare. Psychometric techniques concerned with the assessment of personality may provide a suitable basis for scientific study in this field. Potentially methodologies include: behavioural tests, objective behaviour measures or the quantification of reports from those familiar with the subjects. The assessment of the validity and reliability of the variables measured in these tests is an integral part of their development. Interobserver correlation in an experiment based on the subjective rating of 20 horses with respect to 14 familiar terms used to describe horse personality was generally low. This suggests that, with the exception of the terms 'flighty' and 'sharp', the empirical terminology commonly used to describe horse personality is unreliable.