Translocating populations of endangered animals to predator-free offshore islands is an extreme, and potentially risky, conservation management strategy. There are many examples of failures and successes, however relatively little work has been done to formally quantify the required numbers of translocated individuals to increase the chance of successful establishment. Cook Strait giant weta, Deinacrida rugosa Buller 1870 (Orthoptera: Anostostomatidae), of New Zealand are an example of an at-risk species where extreme conservation measures are needed, due to their ground-dwelling lifestyle making them particularly prone to habitat loss and predation by alien species following human settlement. Here, we have used microsatellite markers to characterise genetic diversity within and between two parental populations (Mana and Stephens Islands), and two translocated populations (Maud and Matiu/Somes Islands), and estimate probabilities of genetic bottlenecks. We have modelled various demographic scenarios to quantify the severity of effective population size fluctuations resulting from translocations using DIYABC, and explored the number of founders needed to retain rare alleles using ALLELERETAIN. Our results suggest that populations have expanded on both Maud Island and Matiu/Somes Island from effective population sizes at time of translocation of 36 and 47 respectively, but provide evidence for a genetic bottleneck on Maud Island. We also show that translocations have had minimal impact on parental effective population sizes. The impact of our results on future conservation management efforts is discussed.