Abstract
If the objective of identifying accident blackspots is to find suitable sites to implement the most- effective remedial works, then blackspot identification techniques should be evaluated in terms of how effectively they fulfil this objective. From this perspective various studies have indicated that accident totals is the most effective criterion to use to select rural accident sites, and that accident rate after total is the most effective criterion for urban intersection site selection. However, other criteria such as number of accidents with common factors or the potential for accident reduction criterion have not yet been evaluated. An improved method to evaluate accident blackspot identification techniques is described.