Driving a car is a complex psychomotor and perceptual task which is subject to impairment by many factors. Several workers have studied the potential effects of drugs and alchol in crash production by epidemiological and laboratory studies. Both types of studies have yielded useful data but their limitations must be borne in mind when applying the results in pratice. Alcohol is obviously the most common single cause of traffic accidents. A progessively increased risk with increasing blood alcohol levels is well documented; fatigue and/or drugs increase this risk. Drugs are related much more infrequently to traffic accidents although on the basis of statistics, there is a potential risk with drug use. However, drugs alone are not as important as alcohol. The most significant drugs as regards driving risk are obviously certain antianxiety agents, hypnotics, stimulants, hallucinogens, marihuana, lithium and narcotic analgesics, as well as ganglionic blocking agents, insulin and sulphonylurea derivates. Patients should not drive after taking these drug until they are objectively fully alert and capable. Anticholinergics, antihistamines, antidepressants, antipsychotics, phenybutazone, indomethacin, alpha-methyldopa, and beta-blockers may in some cases cause central side effects (e.g. drowsiness) strong enough to affect driving performance. After starting therapy with these drugs, or after a significant change in dose, driving should be avoided until it is known that unwanted effects do not occur. Psychotropic drugs may enhance the deleterious effect of alcohol, and with most hypnotics there is still an effect the next morning. Some drugs (e.g. anticonvulsants or antiparkinsonian drugs) may make driving safer, but the disease (epilepsy, Parkinsonism, cardiovascular diseases, psychic disorders, etc.) ofter precludes driving. Clinicians should warn their patients about an impairment of driving skills if this is likely to occur due to the drug or the illness concerned.