This work examines the influence of clutter when presenting information with a head-mounted display (HMD). We compare clutter costs when displays overlay a real-world scene to the costs of visual scanning required when displays are presented separately. Using an HMD in safety-critical environments reduces repetitive visual scanning and head movements that can become effortful with separate displays, such as a tablet. However, a trade-off occurs with overlay displays when low visibility information in the scene is needed or when perceiving text and symbols on the display requires high visual acuity. To examine this scan-clutter tradeoff, participants performed tasks requiring focused attention on either the scene or the display. The HMD either overlaid the critical aspects of the scene or was presented adjacent to the scene. The amount of clutter in both domains was quantified and manipulated. The HMD overlay and adjacent conditions showed similar performance for accuracy, but the overlay condition hindered tasks requiring focused attention on the scene. Perceiving clutter as perceptually closer was attributed to a biological tendency to prioritize information closer to the observer, which disproportionately harmed attention to scene information. Increasing clutter in both domains caused an increasing cost to both speed and accuracy. The results speak favorably to using an HMD, but signal the need to be cautious of the negative effects of clutter in either domain. These results highlight the importance of carefully designing HMDs to minimize clutter, especially when scene information is required.