Abstract
The findings of 28 psychiatric epidemiological studies con ducted in Africa since the late 1960's are briefly reviewed, and problems with the use of standardised psychiatric epidemio logical instruments in Africa are explored with reference to these studies. Regardless of whether the purpose of the inves tigation is to enhance theory development or to provide service providers and policy-makers with accurate estimates of the nature and extent of psychiatric illness, threats to the validity of the whole epidemiological investigation must be evaluated and dealt with. These threats pertain to not only the instru ment, but also the procedures, the study design, and the sampling strategy and sample size. Given human and other re source constraints, psychiatric epidemiologists in Africa will also need to consider using other strategies, including key in formant interviews and indicators of key mental health problems.