Because fruiting trees are uncommon in tropical forests, frugivorous primates experience selective pressure to incorporate knowledge of where to find feeding trees, what to expect when they arrive there, and when they can return after depleting a tree. I investigated these abilities in wild spider monkeys (Ateles belzebuth) in Yasuní, Ecuador, by analyzing the characteristics of feeding trees that drive foraging decisions. Foraging data were derived from four 2-week follows of focal adult females, conducted between May and December 1999, during which I measured and mapped all trees in which the focal subject fed, feeding bout duration, and the number of conspecifics feeding simultaneously with the focal. Taking into account the order in which feeding trees were visited across each follow, I analyzed each foraging decision from the second week of a follow, treating all previously visited trees as options for visits. I scored each option tree in terms of nine ecological variables, including the distance from the decision to each location tree, DBH, recent feeding time and mean feeding times for the focal and other monkeys present, and the interval in hours between the foraging decision and the most recent visit to each option tree. I then examined the predictive strength of the model using logistic regression analysis, comparing characteristics of selected trees to those not selected. The overall model successfully predicted trees selected by focal monkeys (r2 = 0.27). Monkeys preferentially moved to nearby, large canopy trees, in which previous feeding success was high, and which were visited after an interval of 3.5 days. Interval mattered most for medium and large trees, but did not predict selection for trees <10 cm DBH. Despite the large home range and large numbers of trees, Yasuní spider monkeys appeared to integrate spatial, value, and temporal information when deciding where to feed. Am. J. Primatol. 76:1185–1195, 2014. © 2014 Wiley Periodicals, Inc.