A field accident research study was conducted in the City of New Orleans in 1975-76 in order to (1) determine the percentage and relative risk of alcohol involvement in adult pedestrian fatal and injury accidents;​ (2) identify in the alcohol involved accidents any unique accident types, behavioral errors, or other characteristics which distinguish these accidents from non-alcohol pedestrian accidents;​ and (3) study the alcohol use patterns and the drinker classifications of these pedestrians.

The study was based on extensive data collection on adult (age≥14) pedestrian fatal and non-fatal accident cases and the establishment of control groups based on accident and random site sampling. Specifically, the following groups were developed:​

• Fatal Accident Group (N=86) - all pedestrians (ages≥14) who died within 24 hours of a motor vehicle accident in New Orleans during 1972-76.
• Injury (non-fatal) Accident Group (N=180) - pedestrians (ages≥14) taken to Charity Hospital in New Orleans following an accident that occurred between March 1975 and March 1976.
• Accident Site Control Groups (N=1208) - similarly exposed but non-accident involved pedestrians (ages≥14) passing the site of the fatal and injury accidents within ± 30 minutes on the same day of the week as the original accidents.

Random Site Control Group (N=80) - 112 street locations chosen at random and sampled for one hour periods with sampling evenly distributed insofar as possible by time of day and day of week. Positive blood alcohol concentrations (BACs) were found in approximately half of both the fatally injured pedestrians and non-fatally injured group. The outcome indicates that the degree of alcohol involvement in non-fatal crashes (where the pedestrian is treated at a hospital) is as great as in fatal accidents and that a startling proportion of the victims in both groups have very high BACs.

With regards to the relative risk comparisons, three subgroups of control subjects were employed:​ (a) site matched controls, (b) age/sex/site matched controls, and (c) random controls.

Depending upon the control group used, the relative risk of accident involvement increases at BAC levels ≥.10 % and accelerates rapidly at higher BAC levels. The implications of the results are discussed. An analysis of accident types, behavioral errors and predisposing factors surprisingly did not indicate major significant differences between alcohol involved and non-alcohol involved cases. There were, however, indications that (1) the path the pedestrian chose in crossing the road, (2) the culpability of the pedestrian, and (3) the frequency of the pedestrian striking the vehicle were related to the pedestrian's BAC.

Countermeasure implications of these and other results are discussed as well as recommendations for continuing accident research.