Genetic structure and natal origins of green turtle mixed stocks in southern Brazil were assessed based on analyses of mtDNA control region sequences from the Arvoredo Island (n = 115) and Cassino Beach (n = 101) feeding areas. These were compared to other mixed aggregations to examine structuring, and to Atlantic Ocean nesting colonies to evaluate natal origins through Bayesian mixed stock analysis (MSA). In order to develop novel priors, surface drifter trajectories in the Atlantic were analyzed and combined with rookery data, and we used KulbackLeibler information measures in order to compare the difference of information among the 4 proposed priors. Each study area presented 12 haplotypes, 10 of which were shared at similar frequencies. Haplotypes CM-A8 and CM-A5 represented ~60 and 20%, respectively, and remaining haplotypes accounted for < 5% of samples. The 2 study areas were genetically similar to all feeding grounds in the western South Atlantic except Almofala, in northeast Brazil, and genetically different from Caribbean and North American mixed stocks. Drifter trajectory analysis revealed that drifters from Ascension and Trindade Islands have a larger chance of reaching Brazil. The priors drifter data and rookery size/drifter data combined contained the most information, but stock estimates were not greatly changed. MSA indicated that Ascension, Aves/Surinam and Trindade were the main stock contributors to the study areas. Since impacts on mixed stocks may affect populations thousands of km away, the results presented here have important implications for the conservation of this endangered species.KEY WORDS: Green turtle · Feeding grounds · South Brazil · Genetic structure · Natal origins · Dispersal patternsResale or republication not permitted without written consent of the publisher