In the present study, we used a synthetic task environment representing an aerial wilderness search and rescue task to test and validate neurocomputational models of the human hippocampus. Participants completed two tasks: a search task, in which they searched a wooded area for multiple targets, and a route choice task, in which they searched high and low density probability regions, parametrically varied by relative distance, for a single target. The decision task was designed to examine participant evaluation of cost versus expected reward. We compared participant behavior during this task with (1) an optimal model, which makes decisions based on a cost to reward ratio, and (2) a neurocomputational model based on hippocampal architecture, which generates activation from probability weighted reward sites. We discuss initial results using this paradigm and the bases by which we will model performance during unmanned aerial search for multiple targets.