Each year approximately 800 school-aged children are killed in motor vehicle crashes during normal school travel hours. This figure represents about 14 percent of the 5,600 child deaths that occur annually on U.S. roadways and 2 percent of the nation’s yearly total of 40,000 motor vehicle deaths. Of these 800 deaths, about 20 (2 percent)—5 school bus passengers and 15 pedestrians—are school bus–related. The other 98 percent of school-aged deaths occur in passenger vehicles or to pedestrians, bicyclists, or motorcyclists. A disproportionate share of these passenger vehicle–related deaths (approximately 450 of the 800 deaths, or 55 percent) occur when a teenager is driving.
At the same time, approximately 152,000 school-age children are nonfatally injured during normal school travel hours each year. More than 80 percent (about 130,000) of these nonfatal injuries occur in passenger vehicles; only 4 percent (about 6,000) are school bus–related (about 5,500 school bus passengers and 500 school bus pedestrians), 11 percent (about 16,500) occur to pedestrians and bicyclists, and fewer than 1 percent (500) are to passengers in other buses.
When school travel modes are compared, the distribution of injuries and fatalities is found to be quite different from that of trips and miles traveled. Three modes (school buses, other buses, and passenger vehicles with adult drivers) have injury estimates and fatality counts below those expected on the basis of the exposure to risk implied by the number of trips taken or student-miles traveled. For example, school buses represent 25 percent of the miles traveled by students but account for less than 4 percent of the injuries and 2 percent of the fatalities. Conversely, the other three modal classifications (passenger vehicles with teen drivers, bicycling, and walking) have estimated injury rates and fatality counts disproportionately greater than expected on the basis of exposure data. For example, passenger vehicles with teen drivers account for more than half of the injuries and fatalities, a much greater proportion than the 14–16 percent that would be expected on the basis of student-miles and trips.

http://www.schoolsecurity.org/trends/school_violence.html