While their importance in the market-place is steadily increasing in developed (mainly continental Europe) and even in developing countries, compounds included in the broad category of 'cerebroactive' drugs hardly rate a mention in reference pharmacology and therapeutics textbooks. It is an undeniable fact, however, that the principal users or targets of this drug class, mainly elderly people, represent an increasingly worrying problem, with their often puzzling cohort of ill-definable and even less predictable neurological and mental symptoms. The combination of the above factors cannot but produce a rather confused situation, in which the pressure to treat and the adherence to scientifically rigorous assessment are likely to prevail alternately, on a purely casual basis. This review aims to provide sound methodological guidelines for assessment of 'cerebroactive' drugs in a not always easily accessible literature. It covers firstly the general problems of stroke, dementia and 'common symptoms' of the elderly, and then looks in detail at those compounds which have to date attracted most attention (ergot derivatives, cinnarizine, flunarizine, vincamine, eburnamonine, naftidrofuryl, oxpentifylline, piracetam and citicoline), as well as those which are currently considered investigational (choline and lecithin). The pharmacology and available clinical studies of each drug are examined. No therapeutic indication can be derived from the available evidence, as the few positive results do not go beyond random improvement of symptoms. More fundamentally, the lines of research which need to be pursued most intensively relate to better preliminary definition of diagnostic and prognostic criteria and, with the establishment of adequate testing tools for the assessment of behaviour and neuropsychological performance, those basal conditions which are modified 'naturally' or by drugs.