Health policy in the United States has paid scant notice to physical activity until recently. This current policy focus on physical activity has revealed not only that there is less than adequate data about it but also that the single survey questions used for the purpose of measuring its prevalence should be interpreted carefully. A case in point is the example presented in this article, which gives estimates of physical activity prevalence levels for women of child-bearing ages from several National Center for Health Statistics (NCHS) surveys. The amount of physical activity reported, and thus the proportion of women at risk for various diseases due to limited activity, depends on how the question is asked and the type of responses offered as options. Various questions used in three different NCHS surveys produced prevalence estimates of limited physical activity levels ranging from 3.9 to 39.1%. These findings have important implications for survey assessments of physical activity at all policy levels. © 1987.