The increasing use of head-mounted displays (HMDs) for virtual and augmented reality (VR/AR) has raised concerns about the musculoskeletal effects and potential risks of pain and discomfort in the cervical spine. This study aims to determine the differences in loading during fast and slow head movements associated with headsets that have different display masses and inertias typical of VR/AR headsets by examining joint torques, muscle activation, neck posture, and self-reported discomfort scores. Twenty participants completed a series of five movements (slow and fast flexion/extension, slow and fast rotation, and a search task) while wearing thirteen different headset configurations of varying mass and inertia (mass distribution). A custom-designed prototype was used for nine of the thirteen configurations to test three different design parameters: display mass, display mass distance, and counterweights. The remaining four configurations consisted of three commercial headsets (Glass Enterprise 2, Magic Leap 1, and HTC Vive Original) and a no headset control. Joint torques were calculated from inverse dynamics based on motion data of the head and torso. Muscle activity was derived from the EMG signals produced by muscles in the neck. Neck posture, calculated as vertical neck angle, was a measure of how far forward the head was from the seventh cervical vertebra (C7). Increasing display mass led to an increase in cervical spine joint torques, a decrease in vertical neck flexion angle, and an increase in self-reported discomfort. Increasing display mass distance had the same effect. However, joint torques only increased around the occiput-first cervical vertebra (OC1), not around C7. Adding a counterweight decreased the torque around OC1. However, it led to both increases and decreases around C7 depending on the axis of rotation. Counterweights also increased vertical neck flexion angle. Commercial headsets followed the same trends with headset display mass being the key determinant of cervical spine loading. Based on these results, reducing display mass and display mass distance were the most effective ways to minimize cervical spine loading during head movements.