A new experimental model was developed to induce diffuse brain injury (DBI) in rats through pure coronal plane angular acceleration. An impactor was propelled down a guide tube toward the lateral extension of the helmet fixture. Upon impactor-helmet contact, helmet and head were constrained to rotate in the coronal plane. In the present experimental series, the model was optimized to generate rotational kinematics necessary for concussion. Twenty-six rats were subjected to peak angular accelerations of 368 +/- 30 krad/sec2 (mean +/- standard deviation) with 2.1 +/- 0.5-msec durations. Following rotational loading, unconsciousness was defined as time between reversal agent administration and return of corneal reflex. All experimental rats demonstrated transient unconsciousness lasting 8.8 +/- 3.7 min that was significantly longer than control rats. Macroscopic damage was noted in 51% of experimental animals: 38% subarachnoid hemorrhage, and 15% intraparenchymal lesion. Microscopic analysis indicated no evidence of axonal swellings at sacrifice times of 24, 48, 72, and 96 h. All rats survived rotational loading without skull fracture. Injuries were classified as concussion based on transient unconsciousness, scaled biomechanics, limited macroscopic damage, and minimal histological abnormalities. The experimental methodology remains adjustable, permitting investigation of increasing DBI severities through modulation of model parameters, and inclusion of further functional and histological outcome measures.