Purpose of this paper is to demonstrate that the tremendous volume of accident reports prepared on the scene by investigators, not under direct control of an organized research team, is a research resource as yet unexploited to its full potential.

An additional purpose of this paper is to counter a growing sentiment among many informed authorities to reduce, if not virtually eliminate, accident investigation and reporting. It is true that accident information systems are expensive and that it is desirable to get traffic moving after an accident. But, if the data resulting from accident investigation and reporting are valuable research materials, apart from their value in subsequent litigation, it would be undesirable to limit their collection.