The consensus panel agreed that a number of facts must be available in order to establish that the use of a drug results in driving impairment and to justify testing to respond to this hazard. Laboratory studies must be able to produce a dose-related impairment in driving or similar psychomotor skills. Concentrations of the drug and/or its metabolites must be capable of being accurately and quantitatively measured and quantitatively related to the degree of impairment produced. Such impairment should be confirmed in actual highway experience. In addition, simple behavioral tests are needed that can indicate the presence of drug-related impairment to the courts, and that can be incorporated into law as ipso facto evidence of impaired driving. It was the opinion of the panel that such criteria have been met for ethanol, but that it is not certain that they can be met for other drugs that now are of concern for highway safety.