The major sensorimotor deficits that result from traumatic spinal cord injury (SCI) are due to loss of axons in ascending and descending pathways of the white matter (WM). Experimental treatments administered after a standardized SCI can reduce WM loss and long-term functional deficits. Thus, a significant proportion of WM loss occurs secondary to the mechanical injury and may be a target for therapeutic intervention. Presently, we know little of how and when secondary injury mechanisms operate in the WM after SCI. We therefore used a standardized rat model of clinically relevant contusion injury to examine axonal pathology over the first 24 h by light and electron microscopy. Based on qualitative evaluation of tissue at 15 min, 4 h, and 24 h after a "mild" SCI produced with a weight-drop device (10 g x 2.5 cm), we selected areas from the ventromedial WM at the lesion epicenter for quantitative analyses. We compared axon number and the proportion of axons with various axoplasmic and myelin abnormalities over time after SCI, as well as the effect of axon size on degree of pathology and loss. We found by 4 h postinjury (pi) axonal pathology was more severe than at 15 min and that a significant loss of large diameter axons had occurred; no significant additional loss of axons was seen by 24 h pi. When we compared axonal pathology after a more severe contusion (10 g x 17.5 cm), we found a greater loss of axons at 4 h. In addition, a higher proportion of the remaining axons demonstrated pathological alterations. We developed a semi-quantitative Axonal Injury Index (AII) as an overall measure of axonal pathology that was sensitive to the effects of injury severity at 4 h pi. The AII has greater statistical power than our individual measures of axonal pathology. Our results suggest that it may be possible to use the AII at 4 h pi to assess effects of potential therapeutic agents on acute axonal pathology after SCI.