Optimal escape theory seeks to explain variation in the distance to an approaching predator at which the prey initiates escape (flight initiation distance). Flight initiation distance increases when predators pose a greater threat and decreases when escape costs increase. Although optimal escape theory has been highly successful, its predictions have been tested primarily for species that escape to discrete refuges, and most studies have focused on single risk or cost factors. We present data from two experiments in which two risks or a risk and a cost varied in Bonaire whiptail lizards ( Cnemidophorus murinus ) that escaped without entering refuges. Our data verify several predictions about optimal escape for nonrefuging lizard prey. Two risk factors, speed and directness of approach by the predator, interacted. Directly approached lizards had greater flight initiation distances than did indirectly approached lizards when approached rapidly, but shorter flight initiation distances when approached slowly. Flight initiation distance was shorter in the presence of food and during slow versus rapid approaches, but contrary to expectation, food presence and approach speed did not interact. This would be explained if cost curves are nonlinear or if they are parallel rather than intersecting when the predator reaches the prey. More empirical work is needed to determine which risk and cost factors act additively and which act synergistically. The absence of interaction between the risk and cost factors suggests that cost curves were nonlinear.