The effects of degeneration and surgery on cervical spine mechanics are commonly evaluated through in vitro testing and finite element models derived from these tests. The objectives of the current study were to estimate the load applied to the C2 vertebra during in vivo functional flexion-extension and to evaluate the effects of anterior cervical arthrodesis on spine kinetics. Spine and head kinematics from 16 subjects (six arthrodesis patients and ten asymptomatic controls) were determined during functional flexion-extension using dynamic stereo X-ray and conventional reflective markers. Subject-specific inverse dynamics models, including three flexor muscles and four extensor muscles attached to the skull, estimated the force applied to C2. Total force applied to C2 was not significantly different between arthrodesis and control groups at any 10 deg increment of head flexion-extension (all p values ≥ 0.937). Forces applied to C2 were smallest in the neutral position, increased slowly with flexion, and increased rapidly with extension. Muscle moment arms changed significantly during flexion-extension, and were dependent upon the direction of head motion. The results suggest that in vitro protocols and finite element models that apply constant loads to C2 do not accurately represent in vivo cervical spine kinetics.