Optimal-escape theory predicts that flight initiation distance (predator–prey distance when fleeing starts) increases as predation risk increases; optimal-refuge-use theory predicts that time spent hiding in refuge increases as risk upon emergence increases. I tested predictions of an extension of escape theory to a scenario in which an immobile predator stands near an immobile prey. When a previously immobile predator moves, the probability of fleeing is predicted to increase as predation risk increases. This scenario corresponds to ambush. I also examined potential effects of three poorly understood risk factors on time spent hiding in refuge. In the lizard Sceloporus virgatus, probability of fleeing was greater when the predator turned toward rather than away from prey, stood closer to prey before moving, or was oriented parallel to the prey rather than facing in the opposite direction. These findings and previous work indicate that cost–benefit considerations govern responses to ambush predators as well as to openly approaching predators, extending escape theory to encounters in which both predator and prey are initially immobile. Hiding time was unaffected by starting distance (predator–prey distance when approach begins), was longer after the second of two successive approaches, and was shorter early in the day when thermal cost of refuge use was high than later when refuge temperatures were warmer. The positive correlation between first and second hiding times adds to evidence for a boldness syndrome in S. virgatus.