The hypothesis that muscle tension protects the spine from injuries in helicopter scenarios was tested using a finite-element model of the human head and neck. It was compared with cadaver crash sled experiment with good correlation. Then, simulations were performed with a sinusoidal velocity (5-22 G) applied at T1 60 to the horizontal plane. The model with relaxed muscle activation had delayed and decreased peak head rotation compared with passive properties only. Full muscle activation decreased the injury risk for the 13.5-22 G impacts. A sensitivity study of the impact angle showed a very slight variation of the resulting neck flexion, and 1 change affected all ligament injury predictions less than 4%. Finally, simulations with helmets resulted in increased ligament and disc strains with increasing helmet mass and with an anterior or inferior shift of the centre of gravity. It is concluded that the hypothesis seems to hold.