Abstract
Because traffic accidents in residential areas happen in a not enough number for statistical research, it is impossible for a short-range study to use the accident as a measure for unsafety. An other measure for traffic unsafety is the near-miss, or a severe conflict-behaviour between the road users. The possibilities for such conflict-techniques are outlined. A first try of this technique is used in a study in Delft to compare traffic safety in two residential areas, different in lay out.