Juvenile dispersal is a life history strategy found across taxa and habitat types. In this dissertation, I examine juvenile dispersal in an evolutionary context and explore the conditions that support high dispersal, which necessitates subsequent ontogenetic habitat shifts. I use a sea turtle study system to evaluate the genetic and behavioral patterns associated with juvenile dispersal and ontogenetic shifts in marine environments. Specifically, I focus on four sea turtle species found in the northeastern Gulf of Mexico in the early life stage known as the ‘lost years’: green turtles (Chelonia mydas), Kemp’s ridleys (Lepidochelys kempii), loggerheads (Caretta caretta), and hawksbills (Eretmochelys imbricata). I use mitochondrial DNA (mtDNA) haplotypes and ocean transport probabilities to estimate contributions from source rookeries to sampling sites in the Gulf of Mexico as well as assess gaps in the available genetic datasets. The mixed stock analyses indicate that the majority of ‘lost years’ green turtles I sampled originate from nesting beaches along the coast of Mexico, with smaller contributions from Costa Rica and Suriname. More broadly, I identify gaps in available genetic data across species, life stages, and ocean basins, particularly among early juveniles. I use telemetry data to characterize sea turtle ‘lost years’ movements with respect to the continental shelf and proximity to the coast, potential recruitment to coastal habitats, and passive versus active behavior by using oceanographic surface drifters as a control. The West Florida Shelf is a high-use area, particularly among green turtles and Kemp’s ridleys. Finally, I analyze genome-wide single nucleotide polymorphisms to test whether mtDNA haplotype data accurately reflect overall population structure for defining management units. These data provide valuable insight into the elusive ‘lost years’ juvenile dispersal life stage in sea turtles and fill substantial data gaps impeding management for species conservation.