Three sets of studies are surveyed: those providing data on the extent of drinking and driving among youth under 20 years, those that examine alcohol use among youth involved in crashes, and those that estimate the relative risk of a crash for young drinking drivers. This research shows that many 16-to 19-year olds report drinking and driving and some 50 percent of teenagers report riding in a car with a drunken driver. Roadside surveys conducted in Canada and the United States, however, suggest that the lowest frequency of drinking and of high blood alcohol consumption (BAC) levels occurs among young drivers. Nevertheless, the data also reveal that young drinking drivers are more likely to be involved in crashes than older drinking drivers in all BAC ranges. Reasons underlying these statistical findings are explored. One explanation assumes that younger people are inexperienced with drinking and driving, while other researchers believe the comparatively higher crash risk is due to a subset of young drinking drivers who engaged in risky driving behavior and also happen to use alcohol. Implications of these findings for research and prevention programs are discussed. Tables and over 60 references. (Author abstract modified)