Injuries are the leading causes of death for non-infant children and adolescents. Motor vehicles, drownings, fires, firearms, and poisons, in that order, are the major categories of fatal injury for the child population as a whole. Recognition that the agents of injury are the major forms of energy (mechanical, thermal, chemical) has led to an increased focus on the characteristics of energy and factors in children's environments that increase exposure to these agents. A wide range of strategies is available to modify such agents and environments. These strategies would reduce the incidence and severity of injuries, including prevention of creation or release of the agents; separation of the agents from children in time, space, or with physical barriers; and modification of qualities of agents, vehicles and hosts. Research relevant to choice of strategies suggest that changes in injurious agents or environments that do not require modification in behavior of those at risk are usually the most effective and less often generate objections from the general public. However, resistance to regulation on economic, political, and ideological grounds may delay or foreclose the use of a given strategy. Therefore, attempts at injury control should be researched in relatively small-scale experiments before being adopted on a large scale. Such experiments could rule out the programs that have unanticipated harmful effects or no effects, and give more credence in the economic, social and political arenas to those programs that are effective. (Author/MP)