The Office of Statistics and Analysis (OSA), National Highway Traffic Safety Administra- tion (NHTSA), has embarked on a 2-year National Crash Severity Study (NCSS). NCSS is the bridge between the Restraint Systems Evaluation Project, conducted by OSA during 1974-75, and the National Accident Sampling System, scheduled for full implementation in 1980.

The primary objective of NCSS is to provide a nationally representative data base to determine statistical relationships between crash conditions and injury severity and to estimate the distribution of crash conditions among the nation's automobile towaway accidents. Special attention will be devoted to estimating the distribution of delta-V (ΔV, velocity change dwing impact) and the extent to which ΔV can be used to predict injury severity. NCSS will also describe, for each injury with an Abbreviated Injury Scale (AIS) rating equal to or greater than 2 (AIS ≥ 2 includes moderate or more severe injuries), the details of the specific injury, the contact point causing it, the medical treatment required, and the days of disability and work-days lost. These data will be used to support major new analyses on injury causation and injury costs and consequences.

The NCSS data are "nationally representative" in the sense that they are collected in a purposive sample of eight ereas. The NCSS areas have almost the same distribution of central city, suburban, small-town, and rural population as the nation, and there is at least one NCSS area in each of the nation's four demographic regions. Within each NCSS area, accidents are chosen for investigation by strict adherence to a stratified probability sampling scheme. The NCSS sample will eventually contain 25 000 occupants, including approximately 600 fatalities and 1 500 severe (AIS-3,4,5) injuries. This sample will be representative of approximately 100 000 occupants of automobiles towed from accidents in the eight areas.

The NCSS data are collected by seven multidisciplinary accident investigation teams. In each accident, they will obtain only those data needed elements for NCSS analyses. These include measurements of damage and post-crash trajectory required for calculation of ΔV by computer reconstruction of the accident.

A quality control contractor will be responsible for checking that all teams interpret data elements in a consistent manner, that teams adhere strictly to the sampling protocol, and that data completeness and accuracy are maximized.