At the present time, research indicates that the use of a psychotropic drug, such as a tranquilizing agent, an antidepressant, a sedative, a stimulant, a narcotic analgesic, or a drug that modifies perception (e.g., cannabis or a hallucinogen) appears to increase the risk of traffic accident involvement. From previous studies, the drugs most likely to be found in impaired drivers, often combined with alcohol, are sedatives and hypnotics--especially barbiturates and tranquillizers, including diazepam and other benzodiazepine derivatives. Although evidence exists that cannabis impairs driving skills, the absence of a fixed analytical method for the identification of the drug in body fluids has hindered the detection of drivers actually affected by the cannabinoids. An examination of the effects of various drugs upon simulated driving tasks, as well as other tests measuring the effects of various drug dosages upon experimental subjects, note that determination of drug levels in blood that adversely influence driving tasks must be considered in conjunction with a host of factors, including case circumstances, time interval since dosage, the pattern of drug administration, age, food, impaired liver function, and the simultaneous administration of another drug. It is concluded that the practical problems of analysis in relation to traffic safety and restrictive legislation of drug use directed against drivers are similar to other types of toxicological work, in that experimental methods must be specific, sensitive, precise, and under continuous review. However, because of variations between individuals, the concentration of a drug or its metabolite in the blood should not provide the sole evidence of impaired performance, but should be considered in conjunction with the results of a clinical examination. Thirty-five references are provided.