A series of 16 sled impact tests was conducted at the Highway Safety Research Institute sled facility to evaluate the effectiveness of restraint devices and systems currently being used to transport school-bus and wheelchair-seated handicapped children. A sled impact pulse of 20 m.p.h. and 16 G's was used for all tests. Eight tests involved wheelchairs in forward-facing and side-facing orientations for head-on and 33-degree oblique impacts. Another eight tests involved forward-facing bus seats for head-on and 33-degree oblique impacts. The results generally point out the ineffectiveness of many currently used devices and systems for protecting the child in a bus collision. In six of the eight bus seat tests the dummy's head struck the back of the bus seat in front. This was primarily because of a lack of upper-torso restraint. A padded belt commonly used for restraining children in wheelchairs is also inadequate by itself and should only be used with additional thorax and pelvic restraint. The practice of placing wheelchairs in a side-facing orientation was found to be a poor one for the protection of the child.