The prevalence of a variety of risk factors and their strength of asso
ciation with a disease can vary greatly among apparently similar commu
nities. In small communities, risk estimates can also vary from year t
o year. An identification of important risk factors in each community
is then needed so that interventions can be specifically oriented towa
rds the needs of each specific community. The attributable risk is the
adequate measure of association for these purposes. The purpose of th
is paper is to determine the minimum sample size required to detect a
given attributable risk in cross-sectional studies. A table was constr
ucted, presenting the number of exposed subjects necessary to detect a
given attributable risk for different combinations of prevalence of d
isease and prevalence of exposure to a given risk factor, with a power
of 0.80 and alpha of 0.05.