Drug-induced behavioural changes may, over time, be influenced also by metabolic adaptations to the substance and by psychological adaptations, including learning to adjust to the primary behavioural consequences of the drug. For these reasons, the subjects examined to show psychomotor effects should be relatively inexperienced with the class of drug under scrutiny. Thus, immediate psychomotor changes, for example, changes during the first day of medication, are assumed to be less subject to ambiguous interpretation than behaviour assessed after repeated administrations of the drug, for example, over a course of treatment. It is proposed, therefore, that optimal subjects for assessing the psychomotor effects per se of benzodiazepines would be normal volunteers assessed in a manner to reflect initial response to the substance.The design of such inquiries should provide for the standard comparison between premedication behavioural assessments and assessments at fixed intervals during a standard period, for example