Adults with putative attention deficit hyperactivity disorder (ADHD) are increasingly being referred to psychiatric clinics, often following a self-diagnosis, and demanding a prescription for stimulant medication. This has disconcerted many clinicians and started a debate on the appropriateness of this diagnosis in adults (Shaffer, 1994 ;Lomas, 1995 ;Diller, 1996) that is reminiscent of the controversies of the childhood diagnosis in previous years (Lancet, 1986). At issue is not only concern about the widespread use of stimulant medication, but also a neurobiological understanding of impulsivity, hyperactivity and antisocial behaviour and the genesis of some psychiatric disorders in adults. How is the validity of this disorder in adults then to be established ?First, it is useful to remind ourselves that diagnoses are concepts that evolve over time and their usefulness is judged by their ability to inform us about pathophysiology, treatment, possible prevention and prognosis. In psychiatry, most diagnoses continue to be based on symptomatology and course of illness, and their validity must be determined in the absence of gold standards. While construct and concurrent validity are important, predictive validity in terms of treatment and prognosis is pre-eminent in most diagnostic determinations (Kendell, 1989). Furthermore, the tension between categorical and dimensional approaches to classification has not eased completely, with clinicians generally finding the categorical approach more appealing. These principles apply to adult ADHD as much as to other psychiatric syndromes.Secondly, a seemingly simple argument ! If the diagnosis of ADHD in childhood can be unequivocally established, and longitudinal studies can demonstrate the continuation of the disorder into adulthood, the status of the adult disorder would not be difficult to validate. The alternatives are that either the disorder remits in adolescence in all cases, or the adult manifestations are categorically distinct from the childhood manifestations such that a different diagnostic label is appropriate.Controversy exists in relation to its prevalence and its operational criteria, but childhood ADHD is firmly established as a diagnostic entity in child psychiatry clinics. In the United States, as many as 50 % of clinic attendees are given the diagnosis of ADHD (Cantwell, 1996), and prevalence rates are estimated at 3-5 % in the general population (American Psychiatric Association, 1994). The diagnosis is used much less frequently in the United Kingdom (Hoare, 1993) and Australia (Rey & Hutchins, 1993) and this appears to reflect differing diagnostic criteria being used. The ICD-10 (World Health Organization, 1992) has adopted a more restricted definition of the ' hyperkinetic syndrome ' than DSM-IV. No classificatory system has, however, dismissed the diagnosis outright. When diagnostic criteria are used consistently, cross-national differences in prevalence seem to disappear, as evidenced by two recent epidemiological studies in children. A Tennessee, ...