A double-blind controlled comparison of four commonly-used tranquillizing drugs (haloperidol, amylobarbitone sodium, chlordiazepoxide, and trifluoperazine) against placebo was made in their effects on the performance of volunteers during three low speed vehicle-handling tests. The drugs (with the exception of haloperidol) significantly altered driving behaviour though they did not seem to interact significantly with alcohol. There is, therefore, a strong possibility that such drugs will similarly alter driving performance in patients taking them for therapeutic purposes. Since, as these experiments also show, those affected may be subjectively unaware of it, and routine clinical screening is not sensitive enough to detect them, physicians should warn patients of the probability that their driving performance will be affected by such drugs, particularly during the first few days that they are taken.