The study of animal personality variation promises to provide significant new insight into the way that behaviour evolves in animals, along with its ecological and evolutionary influences. We strongly advocate more empirical work in this exciting and rapidly expanding research area, but hope that new studies adopt a more hypothesis-driven and/or experimental approach than seems to be usual at the moment. Here we outline what we feel is "good practice" to the many empiricists that are keen on pursuing work in this field. We highlight the substantial body of theoretical work that exists for providing well-reasoned hypotheses, which new empirical studies should be designed to test. Furthermore, using a brief review of existing work on the behavioural ecology of animal personality variation in the zebra finch - one of the more widely used model systems in this field - we stress the importance of understanding the ecology of the chosen study animal, and the problems that are likely to arise by neglecting to identify or account for the structure of behavioural variation that is often likely to occur.