In 1943, Holbourn proposed that brain injury due to rotations is proportional to angular velocity for short duration impacts and is proportional to angular acceleration for long duration impacts. Gabler et al. (2016) concluded that angular velocity was sufficient to predict brain injury for impacts of up to 100 ms. The objective of this study was to:​ (1) evaluate the correlation of a brain injury metric with angular velocity and angular acceleration for short (<100 ms) and long (>100 ms) duration crash events using idealised pulses;​ and (2) analyse angular velocity and acceleration data from fleet/crash tests and evaluate their feasibility for use in brain injury metric formulation. The study with idealised data demonstrated that both angular velocity and acceleration correlate with brain injury metric up to 100 ms. For crash test data, SAE filtering class CFC 60 resulted in greater peak reduction for angular acceleration (up to 90%), while peak reduction for angular velocity was lower (up to 8%). In addition, collinearity between angular acceleration and velocity was observed when angular accelerations were filtered at CFC 60. Finally, when using crash test data, only angular velocity may be appropriate to use for brain injury formulation.