We review the evidence for behavioral avoidance of plant secondary metabolites (PSMs) and identify how, and the circumstances under which it occurs. Behavioral strategies of avoidance of PSM can only be fully understood in relation to the underlying physiological processes or constraints. There is considerable evidence that animals learn to avoid PSMs on the basis of negative postingestive effects. The extent to which this process determines foraging choices is limited by the ability of animals to experience the consequences of their behaviors and associate particular cues in foods with their specific effects in the body. The proposed learning mechanisms require that animals must at least "sample" plants that contain PSMs. They do not completely avoid PSMs, but there is evidence that they restrict their ingestion to within limits that they are physiologically able to tolerate, and that the amounts of PSM ingested result from a balance between toxicological considerations and the nutrient content of the plant material. These limits are influenced by the kinetics of PSM elimination, which underlies patterns of bite and patch selection from plant parts to landscapes. We suggest that altering spatial location of feeding (to alternative food patches or alternative foods within patches, including plant parts), and temporal distribution of feeding activity, by either cessation of feeding or by continuing to feed, but on alternative foods, can both lead to reduction of the intake and toxic effects of PSMs. We propose that the strategy of avoidance or reduction of intake of PSMs coevolved with the animal's ability to physiologically tolerate their ingestion, and that avoidance and tolerance are inversely related (the avoidance-tolerance continuum). The animals' propensity and ability to seek alternatives also vary with the dispersion of their food resources. Further work is required to test these proposals and integrate temporal and spatial aspects of foraging behavior and its nutritional consequences in relation to PSMs.