Two experiments examined the role that depth plays in the formation of associations during contextual cuing of visual search. Current associative models make predictions about the spatial constraints placed on learning within two-dimensional procedures, but there exists very little evidence of how these predictions translate to three-dimensional space. A virtual reality procedure was used to project the stimuli in three-dimensions. Experiment 1 established a contextual cuing effect using this procedure, while Experiment 2 examined whether the relative distance between repeated distractors and the target, or the position of the distractors relative to the observer modulated contextual cuing. It was found that the contextual cuing effect was consistent across these different conditions. As a result, there was no evidence to suggest that depth information forms a significant part of the representations that form during contextual cuing. These data are therefore broadly consistent with the mechanisms of current associative models of contextual cuing.