Foraging is a complex visual task comprising a mixed array of target and distractor stimuli. Similar to visual search, targets may be distinguished by a single feature or by a conjunction of features. Unlike visual search, foraging arrays contain many targets of multiple types and participants are instructed to select all targets in whatever order they choose. These multiple selections for each trial/array allow for rich analysis of selection sequences and runs that aren’t often possible with other visual tasks. These sequences can also be viewed as a series of individual decisions that are influenced by top-down strategies, bottom up scene features and potentially history of the previous selections. This presentation will show new experimental data and a computational model that tries to break down the critical task features that influence those individual decisions. The experiment data comes from a new ‘snapshot’ foraging task that restricts each foraging selection to its local area to control for global scene statistics. The computational model uses a Bayesian approximation to the Theory of Visual Attention to determine the features that are important for selection in feature and conjunction foraging. Combined, we hope to provide an improved understanding of how attention is allocated in this complex task by treating it as an evolving sequence of selections