Where do people look when searching for multiple targets under time pressure: at salient targets, at locations with high uncertainty about target presence, or somewhere else? Preceding research suggests that people tend to look at salient targets. This is suboptimal, because educated guesses can be made about target presence at these locations without looking (Verghese, 2012). We ran an experiment and constructed Bayesian models to test the generality of this finding. Participants saw stimuli at two locations for 400 msec (i.e., allowing only 1 saccade), and then judged target presence at each location. Noise of low or high contrast was superimposed at the two locations. We observed individual differences in saccade strategies. One participant made no saccades, while achieving reasonable performance. Others applied a mixture of strategies, sometimes favoring salient targets, sometimes favoring uncertain locations. This work provides further insight into task and cognitive constraints that influence saccade strategy selection.