Abstract: Performance of horses, whether in sports or in leisure, depends on both physical abilities as well as temperament. The aim of the present work was to measure individual variation and consistency of behavioural variables, related to temperament, in young horses of the same breed and age, and reared under controlled housing conditions and management. A total of 41 Dutch Warmblood horses were tested at 9, 10, 21 and 22 months of age in two behavioural tests, i.e. the novel object test and the handling test. In the novel object test horses were confronted with an open umbrella that was lowered from the ceiling. In the handling test horses were led by a human to cross a bridge. Per test, behavioural variables in the following behavioural classes were observed: locomotor activity, latency times, postural expressions and vocalisations. Within years, all behavioural variables in the handling test, and all but two in the novel object test were positively correlated (0.36<Rs<0.81, P<0.05). For both tests, at 9, 10, 21 and 22 months of age, a principal component analysis (PCA) was carried out to examine whether there were indications for underlying components of these individual behavioural variables that could possibly serve as measures for temperamental traits. The first component in the novel object test could be regarded as `flightiness' and the second as `sensitiveness'. In the handling test, the first component was suggested to relate to `patience', the second component to `willingness to perform'. The temperamental trait `flightiness' (novel object test) as well as the temperamental trait `patience' (handling test) were positively correlated within both years (0.36<Rs<0.65, P<0.05). For the traits `sensitiveness' (novel object test) and `willingness to perform' (handling test) a positive correlation was only found within the first year (0.44<Rs<0.57, P<0.01). A few individual behavioural variables showed consistency over years. Additionally, just one out of four temperamental traits, namely `flightiness', proved to be consistent over years (Rs=0.49, P<0.01). The temperamental trait `patience' showed a trend between years (Rs=0.31, 0.05<P<0.1). It is concluded that the behavioural tests employed in the present study can be used to reliably identify individual behavioural variables and temperamental traits in young horses. Long-term consistency, i.e. between subsequent years, could not be demonstrated convincingly. Nevertheless, future work may indicate that employing the same approach and considering an even longer time period or different phases of the horse's life, long-term consistency does exist.