Sea turtles are long lived marine reptiles, with most species having long distance, ontogenetic migrations. Movement studies have been biased towards surface and nearshore movements of females. I apply a variety of quantitative approaches to biotelemetry data for multiple species and life stages to address knowledge gaps in the movement ecology of sea turtles.In chapter 1, neonate Caribbean leatherback turtles (Dermochelys coriacea) were acoustically tracked during early dispersal. Movement and statistical models showed that the main external drivers of behavioral differences were abiotic, namely the strength and direction of the currents. Variable behavior may have implications in the form of carry over effects if it results in differential long-term survival. In chapter 2, the dive behavior of adult leatherback turtles in the Eastern Pacific (EP) was investigated using a clustering analysis, convolutional neural network, and statistical model. Results found multiple ecologically distinct vertical behaviors, supporting the use of niche switching in the vertical dimension for EP leatherbacks during their migration. In chapter 3, the vertical behaviors from chapter 2 and horizontal EP leatherback locations were used to create both a novel movement model and a dynamic management tool. Results produced monthly maps and quantitative indices informing bycatch risk, supporting existing conservation efforts for this population. In chapter 4, miniature satellite tags were deployed on post-captive juvenile green turtles (Chelonia mydas) in multiple age classes. Comparisons with current-corrected and passive drift trajectories, a clustering analysis, and a statistical model all revealed significant differences in movement behaviors with current velocities and floating algae presence. Results support the importance of external (abiotic) factors and release effects for differences in juvenile movement. This dissertation addresses key questions on the movement ecology of sea turtles for multiple species and life stages. Throughout, the physical environment is demonstrated to be an external driver of individual variation in movement. Movement behavior is additionally found to vary in multiple dimensions (e.g., horizontal and vertical). Despite individual variations, mean patterns of movement and behavior are identified for each population. The results found have broader implications for both conservation science and movement ecology studies of other long-distance migrants.