A series of experiments were conducted in which the neck muscles of volunteer subjects were dynamically and statistically loaded by systematic variations of twenty-four headgear configurations consisting of eight different centers-of-gravity (CGs) times three different weights. Six subjects would rotate their heads laterally (from side-to-side) for 30 min with each of the headgear loading combinations. Immediately thereafter, the subject would position his head in an isometric head dynamometer and exert a sustained right lateral neck contraction or forward neck contraction at 70% of his maximum strength, during which endurance time (to fatigue) was recorded. The results indicate that the computer model makes reasonable predictions within the boundary conditions. Input data outside the boundary conditions is rejected. The assumption of insensitivity to vertical loading is demonstrated. The assumption of bilateral symmetric response was confirmed for the 1.45 kg and 2.27 kg helmet loads. However, this assumption was not confirmed for the 4.09 kg helmet load. It is concluded from the computer model that aftward, midline loading is the optimal CG location (i.e. maximal endurance) for heavier helmets in the 3–4 km range.