Pending changes in European legislation ban the use of wild-caught animals in research. This change is partly justified on the assumption that captive-breeding (or hand-rearing) increases welfare of captive animals because these practices result in animals with reduced fear of humans. However, there are few actual data on the long-term behavioural effects of captive-breeding in non-domestic species, and these are urgently needed in order to understand the welfare and scientific consequences of adopting this practice. We compared the response of hand-reared and wild-caught starlings to the presence of a human in the laboratory. During human presence, all birds increased their general locomotor activity but the wild-caught birds moved away from the human and were less active than the hand-reared birds. After the human departed, the wild-caught birds were slower to decrease their activity back towards baseline levels, and showed a dramatic increase in time at the periphery of the cage compared with the hand-reared birds. We interpret these data as showing evidence of a greater fear response in wild-caught birds with initial withdrawal followed by a subsequent rebound of prolonged attempts to escape the cage. We found no effects of environmental enrichment. However, birds in cages on low shelves were less active than birds on upper shelves, and showed a greater increase in the time spent at the periphery of their cages after the human departed, perhaps indicating that the lower cages were more stressful. In demonstrating reduced fear of humans in hand-reared birds, our results support one of the proposed welfare benefits of this practice, but without further data on the possible welfare costs of hand-rearing, it is not yet possible to reach a general conclusion about its net welfare impact. However, our results confirm a clear scientific impact of both hand-rearing and cage position at the behavioural level.