Tests of optimal foraging theory have shown that many predators are selective about which prey and which patches should be utilized. I hypothesize that prey species "exploit" this choosiness by evolving characteristics that cause predators to choose alternate prey. Specifically, prey should evolve traits that increase the probability of predator death, decrease the per prey or per patch nutritional intake, increase processing time, and advertise (or mimic advertisements of) undesirable traits. Predator choosiness allows prey to divert the predator instead of defeating it. The evolution of a long-term, prudent foraging strategy requires that three conditions be met: (1) The forager must use resources from a discrete subpopulation; (2) use of that subpopulation must be relatively exclusive; (3) the resource population must respond in such a way that a long-term strategy provides an economic advantage. For the horned lizard-ant system, conditions (1) and (2) were tested by tagging lizards with transmitters or radioactive tags. Horned lizards occupy home ranges much smaller than would be expected if they moved at random and home range overlap was less than expected by random placement of home ranges, thus conditions (1) and (2) were not rejected. Most techniques of home range study do not distinguish random from nonrandom movement. Condition (3) was tested by subjecting ant colonies to various levels of artificial predation. In none of five experiments was the result obtained that an increased harvest intensity led to a decrease in long-term yield; condition (3) is tentatively rejected. Ant colonies shut down in response to predation; this puts a ceiling on their losses. Short-term foraging models were tested for horned lizards foraging at ant colonies. A prediction of the marginal value theorem was not rejected: Horned lizards tended to leave colonies when their instantaneous rate of harvest at that colony had fallen to their average rate of harvest for the day. Another short-term prediction, however, was rejected: Lizards did not stay longer at the "better" of two colonies. A more liberal version of the same prediction was not rejected. Apparently, horned lizards forage adaptively but not optimally.