Most measures of predation risk have evaluated the baseline of risk according to habitat structure, when supplemental food availability was constant. Fewer studies have analyzed the effect of a predator's presence or abundance, and those studies have usually been conducted under controlled conditions or using scats, urines, or odors. We tested if apprehension in collared peccaries (Dicotyles angulatus) was affected not just by habitat structure but also by level of resource availability, and presence of top predators (puma [Puma concolor] and jaguar [Panthera onca]) in the immediate area. We classified microhabitats with different levels of risk based on vegetation structure, and compared apprehension of peccaries in shrubland, edge, and forest with sparse understory. To evaluate perceived predation risk, we used giving up densities (GUDs) in 3 trials with different food quantity. We related daily GUDs to predator presence, according to pictures recorded on trail cameras located around the experimental sites. The shrubland microhabitat was classified as the most dangerous, where peccaries were most fearful, while the edge and the forest interior were considered safer. Visibility was the habitat structure variable that most explained the levels of apprehension. Results from trials on days with the greatest numbers of predator photos were more similar among microhabitats, and peccaries left higher GUDs on the days when predators were photographed in the area. Food availability had a minor influence on GUDs. Sites and microhabitats that favor ambush cover are more risky for prey. Predator presence in the area modified temporally the perception of predation risk in prey, which expressed more apprehension and foraged less in the area at those times.