Patch use under predation risk requires animals to trade off needs for safety and food. I consider within a common framework four ways of modeling how safety and energy influence fitness. These models make similar predictions for when a forager should stop feeding from depletable patches that vary in predation risk. In accord with Brown (1988), patches should be abandoned when the harvest rate no longer compensates for the metabolic, predation, and missed opportunity costs of foraging. In accord with Gilliam & Fraser (1987), safe and risky patches should be left at the same ratio of predation risk to net feeding rate when the alternative activities include resting in a safe refuge to save energy. The quitting harvest rate of the forager within a food patch should be sensitive to predation risk and to the marginal value of food (the fitness value of an additional food item). Quitting harvest rates should be higher in food patches with higher predation risk. Quitting harvest rates and the costs of predation should increase as the marginal value of food decreases. The difference in quitting harvest rates between patches with high and low predation risk should increase as the marginal value of food decreases. Fortunately, the shared qualitative predictions of these models have and can be tested. Unfortunately, distinguishing among the models' differing quantitative predictions shall prove challenging.