The relationship of drugs and traffic crashes has not been defined. Present knowledge suggests that a problem may exist, but research has not established the risk of any drug other than alcohol. Major methodological issues in epidemilogical and experimental research are discussed. Major information needs are defined. Specific research recommendations are presented. The highest research priority is to conduct large-scale field studies to determine the prevalence of drugs in drivers. Until these studies are done, the drug and driving problem will remain an undefined issue.