Control schemes are central to the player experience as they dictate how users interact with the game world. Since the 1990s, desktop 3D games have used a standard control scheme (often called “mouselook”) in which the mouse simultaneously rotates the camera view, aims, and steers the player’s avatar. Virtual reality games require a different approach because head-mounted displays introduce an additional input channel. To date, research has focused on the usability of hardware input devices, while control schemes have received little attention. Furthermore, VR games do not have a standard control scheme as of yet. To address this gap, we conducted a mixed-methods study that compared three types of control schemes for first-person virtual reality games and measured their effect on player experience, performance and immersion. We found that familiar control schemes that coupled control of the camera view and steering the avatar, had no dead zone (i.e., “keyhole” or area of the screen in which input is not recognized), and allowed players to use the mouse to control all in-game functions were preferred and resulted in the best performance and immersion scores.