A modified Marmarou impact acceleration injury model was developed to study the kinematics of the rat head to quantify traumatic axonal injury (TAI) in the corpus callosum (CC) and brainstem pyramidal tract (Py), to determine injury predictors and to establish injury thresholds for severe TAI. Thirty-one anesthetized male Sprague-Dawley rats (392 ± 13 grams) were impacted using a modified impact acceleration injury device from 2.25 m and 1.25 m heights. Beta-amyloid precursor protein (β-APP) immunocytochemistry was used to assess and quantify axonal changes in CC and Py. Over 600 injury maps in CC and Py were constructed in the 31 impacted rats. TAI distribution along the rostro-caudal direction in CC and Py was determined. Linear and angular responses of the rat head were monitored and measured in vivo with an attached accelerometer and angular rate sensor, and were correlated to TAI data. Logistic regression analysis suggested that the occurrence of severe TAI in CC was best predicted by average linear acceleration, followed by power and time to surface righting. The combination of average linear acceleration and time to surface righting showed an improved predictive result. In Py, severe TAI was best predicted by time to surface righting, followed by peak and average angular velocity. When both CC and Py were combined, power was the best predictor, and the combined average linear acceleration and average angular velocity was also found to have good injury predictive ability. Receiver operator characteristic curves were used to assess the predictive power of individual and paired injury predictors. TAI tolerance curves were also proposed in this study.