Various embodiments are provided herein for modifying a virtual scene based on eye-tracking. A computing device coupled to a HMD can provide a virtual scene for display on the HMD. The computing device can receive sensor data from a set of eye-tracking sensors coupled to the HMD. Based on the received sensor data, the computing device can determine a set of focal regions of the displayed virtual scene, including a perifoveal region of the displayed virtual scene. A portion of the virtual scene can then be modified based, in part, on a determination that the portion is outside of the determined perifoveal region.