Abstract
A method of determining what changes to the road in a given area will be most effective in reducing accidents with the limited funds available is discussed. Traffic accident classification, the records required for black-spot identification, and the various aspects to consider when inspecting accident sites are described. Two cost-benefit techniques for determining the priorities of improvements are described in detail, and some estimates of the costs of accidents and improvements given as guidance. A method is described by which the engineer can determine statistically the effect the improvement has had on reducing accidents.