The endangered Mexican fishing bat, Myotis vivesi, appears to have suffered widespread extinction and population decline on islands throughout the Gulf of California, largely due to predation by introduced cats and rats. To restore populations of fishing bats and other native species, conservation efforts have focused on eradicating introduced vertebrates from several Gulf islands. These efforts assume that individuals from existing populations will recolonize islands and that continued dispersal will help sustain vulnerable populations thereafter. However, the extent of inter-island dispersal in fishing bats is unknown. In this study we analyzed patterns of genetic variation to gauge the extent of gene flow and, thus, potential dispersal among islands. DNA was sampled from 257 fishing bats on 11 Gulf islands (separated by ca. 6-685 km of open water), and individuals were genotyped at six microsatellite loci and haplotyped at a 282 bp fragment of the mtDNA control region. With microsatellites, we found weak population genetic structure and a pattern of isolation by distance, while with mtDNA we found strong structure but no isolation by distance. Our results indicate that island subpopulations separated by large expanses of open water are nonetheless capable of maintaining high genetic diversity and high rates of gene flow. Unfortunately, little is known about the spatial patterns of dispersal or mating system of fishing bats, and these behavioral factors, in particular female philopatry, might reduce the probability of the species recolonizing Gulf islands.