Escape from predators by desert iguanas (Dipsosaurus dorsalis) conforms to predictions of optimal escape theory based on risk. I simulated an approaching predator to study risk factors. The primary response variable was approach distance (= flight-initiation distance), i.e., the distance between predator and prey when the prey initiates escape. In additional studies, I recorded whether lizards permitted me to approach close enough to noose them (an indicator of wariness) and the method of escape. Approach distance was greater when the predator approached rapidly than slowly and directly than indirectly, and when the predator turned toward the lizard rather than away. It was greater in open than in more densely covered habitats, which may reflect greater risk due to conspicuousness and (or) a greater distance to refuge. Early in the day at lower air temperatures, desert iguanas permitted a closer approach before initiating escape. While basking after emergence from burrows, lizards escaped into burrows; later in the day they fled. Lizards that fled had high body temperatures; a single individual captured immediately after entering a burrow had a lower body temperature. Lizards presumably enter burrows when low body temperature limits the running speed, but burrow use is costly because attainment of the activity temperature is delayed because of time elapsed and the temperature decrease in burrows.