It is now well understood that climate change has the potential to dramatically affect biodiversity, with effects on spatio-temporal distribution patterns, trophic relationships and survivorship. In the marine turtles, sex is determined by incubation temperature, such that warming temperatures could lead to a higher production of female hatchlings. By measuring nest temperature, and using a model to relate the incubation temperature to sex ratio, we estimate that Caribbean Colombian leatherback sea turtles currently produce approximately 92% female hatchlings. We modelled the relationship between incubation, sand and air temperature, and under all future climate change scenarios (0.4–6.0 °C warming over the next 100 years), complete feminization could occur, as soon as the next decade. However, male producing refugia exist in the periphery of smaller nests (0.7 °C cooler at the bottom than at the centre), within beaches (0.3 °C cooler in the vegetation line and inter-tidal zone) and between beaches (0.4 °C higher on dark beaches), and these natural refugia could be assigned preferential conservation status. However, there exists a need to develop strategies that may ameliorate deleterious effects of climate-induced temperature changes in the future. We experimentally shaded clutches using screening material, and found that it was effective in reducing nest temperature, producing a higher proportion of male hatchlings, without compromising the fitness or hatching success. Artificial shade in hatcheries is a very useful and simple tool in years or periods of high environmental temperatures. Nevertheless, this is only an emergency response to the severe impacts that will eventually have to be reversed if we are to guarantee the stability of the populations.