After outlining the main goals of general population surveys of substance use and abuse, this review discusses practical approaches to problems of measurement, using examples from alcohol research. Problems of validity and comparability are discussed as well. The review shows that there are many problems of measurement and that there are no standards that define "good quality" survey work. Such criteria are required, not only to improve the quality of surveys, but also to attain greater comparability between studies conducted in different countries at different times. One necessary step in this direction would be more extensive and systematic reviews for each substance. Examples of such reviews already exist for alcohol and illicit drug use. There are also few studies that consider the measurement problems of the use of various substances in the same context. As the number of studies is increasing rapidly and at the same time new ideas of survey methodology are emerging, updated reviews would be required at least once in a decade. Thus far, publications that summarize experience from surveys on substance use have been sparse; therefore, it is reasonable to propose that some international agency, such as the World Health Organization, should encourage attempts to review methodological issues in measuring psychoactive substance use on a more regular basis. 42 references