Animal personality has become a major topic in animal behaviour. Much recent progress has come from the use of the open-field test, which is routinely used to separate individuals into fast and slow explorers. However, there is no standard way to conduct the test and it is unclear whether the test really measures exploration. Here, we addressed these issues by combining an open-field test with a battery of independent assays intended to assess the convergent and discriminant validity of the behavioural traits suggested to measure exploration in the open-field test. Our study subjects were common mynas, Acridotheres tristis, introduced to Australia. The analyses confirmed that the open-field test allows individuals to be separated according to their propensity to explore, mainly through metrics related to spatial and object examination of the novel cage. However, other metrics classically used to describe exploratory behaviour, such as the latency to enter the novel space, reflected shyness rather than exploration. The open-field test can therefore be a powerful tool to investigate personality, but only if the biological meanings of the metrics derived from the test are properly validated with independent behavioural assays.