Some of the results formerly included in the dossiers under the heading of 'General Pharmacology: Central Nervous System' were intended to document most of the information now covered by the CNS Safety Pharmacology. This paper allows to point out some general experimental conditions relating to the choice of (1) suitable animal species whose use should be advocated, (2) the route by which the effects of the drugs should be studied and (3) the controls of the experimental context (surroundings, temperature, schedules of the trials, interference with food-intake episodes, influence of noises and smells, expertise of the experimental staff) likely to optimize the predictive value of the trials. Besides the core of investigations recommended by most of the regulatory authorities (changes in gross behaviour, motor coordination, locomotor activity, interaction with hypnotics, convulsant potential), the need to include or not some investigations such as EEG studies and the prediction of adverse events, such as anxiolytic and amnesia potentials, is considered. The place, within the scope of CNS Safety Pharmacology, of studies intended to document the potential for drug abuse and dependence is mentioned and some information given about the methods generally used.