We aimed to verify whether Continuous Qualitative Behaviour Assessment (10 observers used a list of six qualitative descriptors) paired with Temporal Dominant Behavioural Expression (the same observers were asked to select the dominant descriptor and to score its intensity level) was able to monitor fluctuations of animal behaviour expression over time. We applied these techniques to three groups of juvenile goats either weaned (group C), or un-weaned (groups WOM and WM). Each animal was separated from its group, moved to group C and tested for 30 min either while their mothers were at pasture, or while their mothers were in an adjacent pen (group WOM and WM, respectively). Animals from group C were separated from their group and immediately reintroduced to it. TDBE duration and score of each descriptor of behavioural expression were able to detect differences among groups but were unable to describe how the behaviour of the goats changed as the time progressed. TDBE curves described the evolution of each behavioural expression of each animal over time but were unable to detect differences among groups. The χ2 test conducted on peaks of dominance, albeit displaying the variations of the behavioural expression over time and allowing the assessment of differences among groups, focussed on occurrences of higher agreement between observers while neglecting most of the information concerning the descriptors above the level of significance. Conversely, based on mixed analysis of variance with the fixed effects of group, test interval and group x test interval (animal nested into group and observer were considered to be random), most of the descriptors were able to discriminate the three experimental groups while preserving the information on the fluctuations of the behavioural expression of the animals during the test.