Difficulties in collecting reliable data on drug use in the context of an exploratory trial of a drug prevention intervention have led to a consideration of the wider issues of the reliability of routinely collected data. A problem is identified, and although its magnitude cannot currently be described, its nature and possible significance are considered. In this methodological discussion, significant practical and ethical issues are explored, including some of the complexities inherent in conducting research on drug use in schools, along with the need to develop improved methods of study which address the specific challenges posed by the school environment. Implications for the development of effective drug prevention interventions in schools are also examined.