Dispersal drives extinction-recolonization dynamics of metapopulations and is necessary for endangered species to recolonize former ranges. Yet few studies quantify dispersal and even fewer examine consistency of dispersal over many years. The northern elephant seal (Mirounga angustirostris) provides an example of the importance of dispersal. It quickly recolonized its full range after near extirpation by 19th-century hunting, and though dispersal was observed it was not quantified. Here we enumerate lifetime dispersal events among female pups given permanent marks during 1994–2010 at two colonies, then correct for detection biases in estimating bidirectional dispersal rates. An average of 16% of females born at the Piedras Blancas colony dispersed northward 200 km to breed at Año Nuevo, while 8.0% of those born at Año Nuevo dispersed southward to Piedras Blancas. The northward rate fluctuated considerably but was higher than southward in 15 of 17 cohorts. The population at Piedras Blancas expanded 15-fold during the study, while Año Nuevo's changed only slightly, but the expectation that seals would emigrate away from high-density colonies was not supported. During the 1990s, dispersal was higher away from the small colony toward the large. Moreover, cohorts born later at Piedras Blancas, when the colony had grown, dispersed no more than early cohorts. Consistently high natal dispersal in northern elephant seals means the population must be considered a single large unit in terms of response to environmental change. High dispersal was fortuitous to the past recovery of the species, and continued dispersal means elephant seals will likely expand their range further. 

Female elephant seals were marked soon after weaning by inserting plastic cattle tags into their flipper webbing. Numbers on tags identified age and birth place of each individual when observed as a breeding adult. Observations at both colonies were done regularly throughout every breeding season to locate and identify tagged seals. Dispersal was estimated for each birth cohort as the proportion observed breeding at the foreign colony to the total observed breeding.

Data are provided as a tab-delimited table having counts of resident and emigrant females per cohort. See README.md.