Abstract Both the development rate and sex of sea turtle embryos depend on incubation temperature, as all species of sea turtle are ectotherms and show temperature‐dependent sex determination (TSD). Theory predicts that selection should act on populations to optimize developmental times and primary sex ratios. In this study, we use a consistent methodology to measure development rates and model the reaction norm that defines TSD in three populations of flatback turtles Natator depressus and two populations of green turtles Chelonia mydas that nest in Western Australia. We show that development rates vary between and within species, likely reflecting adaptation to local beach microclimates. Similarly, the parameters that define the TSD reaction norm vary between the two species, and among N. depressus populations, with pivotal temperatures (TPIV) varying by 1.5°C (29.6–31.1°C), and the transitional ranges of temperatures (TRT) varying by 1.4–3.3°C. In contrast, we found a similar TPIV for the two C. mydas populations, but a wider TRT at the northernmost tropical rookery. Our findings support the view that thermal parameters in geographically separated populations of sea turtles are broadly similar, but the variation we describe will be highly relevant for predicting how populations will respond to climate change. A free Plain Language Summary can be found within the Supporting Information of this article.