An increase in temperature, within bounds, will accelerate development of reptile embryos, and morphogenesis can be normal over a range of temperatures despite those varying rates of development. Less well understood is the form of the relationship that best describes variation in developmental rate with temperature. In this article, we apply a linear degree.hour model, an empirical curvilinear model, a biophysical model, and a polynomial model to data on rates of embryonic development and temperature in the pig-nosed turtle Carettochelys insculpta from northern Australia. The curvilinear models, which have been applied with success to development of insects, describe the embryonic development of turtles well. When fluctuating temperatures extend beyond the constant temperatures that support successful incubation, the curvilinear models continue to perform well, whereas the linear model predictions fail. Sensitivity analysis indicates that under some circumstances, incubation duration may be increased by diel temperature fluctuations, independent of an influence of mean temperature. In other circumstances, incubation duration may be decreased, and in still other circumstances, diel temperature fluctuations will have no impact on incubation duration. This adds an additional dimension to our understanding of how thermal regimes can be selected or manipulated by reptiles to optimise incubation duration and the timing of offspring emergence.