This review considers the issues involved in measuring the community prevalence of asthma, particularly in the context of international comparisons. We argue that there is no gold standard definition for measuring asthma prevalence, and discuss the currently available methods of case ascertainment. Prevalence studies, if they are to be generalizable, need to involve large sample sizes with high response rates. This necessitates methods that are simple, inexpensive and practicable, but also as sensitive and specific for asthma as possible. We discuss some of the issues that are specific to comparisons of asthma prevalence between diverse populations, and suggest that large surveys using written or video questionnaires of self reported symptoms validated in all of the target populations are the method of choice.