Driver error is the cause of most accidents. Yet "active safety" is difficult to analyze due to the complexity of the processes involved and no unanimity of analytical procedures exists. The purpose of the paper is to determine the investigatory steps necessary for any complete analysis of active safety. These steps are discussed individually and they include:1. Analytical treatment within the parameters of the driver-vehicle-surroundings system; the multiplicity of such treatments suggests that no one approach is universally valid.2. Investigations of driver behavior and vehicle response in driving simulators; such simulators offer the possibility of reproducible, experimental data.3. Determination of driver reaction and vehicle response to driving incidents involving artificially-induced danger and surprise; such tests can safely be conducted on driving courses.4. Reconstruction and evaluation of traffic accidents through use of questionnaires; Daimler-Benz has already made use of such questionnaires. The results are discussed.5. Evaluation of comparable data obtained from drive-recorders; such mechanisms, comparable to those used on airplanes, would be expensive but could provide valuable data.