Abstract
Traffic conflict data have been obtained for six intersections using the technique of observation and definition of conflicts established in a previous study at one of the sites. Serious conflict and injury accident frequencies are shown to be positively related between different junctions and between different locations within a given junction. These data validate the use of the traffic conflict technique as a relatively rapid method of study of junction safety. At each intersection studies of vehicle speed and flow have been made and the important influence on conflicts of vehicles other than those primarily involved has again been shown. (Author/publisher)