A number of microvascular changes, such as the development of astrocyte lucency, increased endothelial pit/vesicle activity, development of crater like lesions, and endothelial microvilli have been reported after injury to the brain. Lateral head acceleration in the non-human primate, however, still provides the best experimental model for human diffuse axonal injury. No attempt has yet been made to document the spatial extent or time course of the microvascular response to acceleration injury to the head. We have examined the brains of baboons 1, 4, 6, and 12 h and 7 days after acceleration injury to the head to analyse the microvascular response. In the experimental animals there was a short-term rise in intracranial pressure followed by a long-term resolution, and a reduction in both mean arterial blood pressure and cerebral perfusion pressure which, however, never dropped below 75 % of baseline for more than 5 min after injury in any animal.We found evidence for extravasation of blood in a small number of blood vessels in all parts of the brain. Interendothelial tight junctions are not disrupted. Pit/vesicle activity rises in the 1st h in the occipital cortex, but not until 4 h in the frontal cortex, and remains elevated for at least 7 days. There is little change in the thalamus. Development of microvilli is most rapid in the frontal cortex with peak values at 1 h, but slower in the thalamus and occipital cortex where peak values are only obtained at 6 h. Highest numbers of microvilli occur in parasagittal regions of the brain. Lucency of the foot processes of perivascular astrocytes develops throughout the diencephalon and telencephalon within 1 h of injury and is most marked in the frontal cortex 6 h after injury. Glycogen deposits occur in astrocytes throughout the brain and are maximal at 12 h. There is, therefore, a differential rate of response of endothelial and astrocyte changes in different parts of the brain after lateral head acceleration.We compare our findings with others in the literature dealing with brain injury. We postulate that lateral acceleration of the head results in widespread, non-disruptive mechanical stress and strain to the brain microvasculature, and suggest that a complex of pathophysiological changes acts throughout the whole brain.