The zebra finch (Taeniopygia guttata) is a small Australian grassland songbird that has been domesticated over the past two centuries. Because it is easy to breed in captivity, it has become a widely used study organism, especially in behavioural research. Most work has been conducted on domesticated populations maintained at numerous laboratories in Europe and North America. However, little is known about the extent to which, during the process of domestication, captive populations have gone through bottlenecks in population size, leading to inbred and potentially genetically differentiated study populations. This is an important issue, because (i) behavioural studies on captive populations might suffer from artefacts arising from high levels of inbreeding or lack of genetic variation in such populations, and (ii) it may hamper the comparability of research findings. To address this issue, we genotyped 1000 zebra finches from 18 captive and two wild populations at 10 highly variable microsatellite loci. We found that all captive populations have lost some of the genetic variability present in the wild, but there is no evidence that they have gone through a severe bottleneck, as the average captive population still showed a mean of 11.7 alleles per locus, compared to a mean of 19.3 alleles/locus for wild zebra finches. We found significant differentiation between the captive populations (F(ST) = 0.062). Patterns of genetic similarity closely match geographical relationships, so the most pronounced differences occur between the three continents: Australia, North America, and Europe. By providing a tree of the genetic similarity of the different captive populations, we hope to contribute to a better understanding of variation in research findings obtained by different laboratories.