Identifying links between breeding and non-breeding sites in migratory animals is an important step in understanding their ecology. Recognising the relative importance of foraging areas and ascertaining site-specific levels of recruitment can provide fundamental and applied insights. Here, satellite telemetry and the stable isotope ratios (δ13C, δ15N and δ34S) of 230 green turtles Chelonia mydas from a regionally important rookery in northern Cyprus were employed to evaluate the relative importance of 4 foraging areas. A preliminary analysis of stable isotope ratios suggested that a major foraging area had been missed through satellite telemetry as a large proportion of turtles had isotope ratios that did not correspond to sites previously identified. Stable isotope ratios were then employed to select 5 turtles to be fitted with platform terminal transmitters in 2015. All 5 turtles were subsequently tracked to the same location, Lake Bardawil in Egypt. Serially collected tissue samples from 45 females, ranging over 2 to 4 breeding seasons, suggested that foraging site fidelity was very common, with 82% of females exhibiting extremely high temporal consistency in isotope ratios. Quantifying fidelity allowed an evaluation of foraging area - specific contributions to each breeding cohort over the past 2 decades and demonstrated that recruitment was unequal among sites, and dynamic over time, with Egypt now currently the major contributor to the nesting aggregation. This work demonstrates the utility of stable isotope analysis to elucidate the spatial ecology of cryptic taxa and illustrates how more robust baselines can be assembled against which to measure the success of future marine conservation initiatives.