OBJECTIVE:This research addresses the most frequently cited methodological criticism of the Epidemiologic Catchment Area (ECA) surveys ‐ which might lead to significantly lower estimates of true prevalence rates for late‐life depression ‐ the inability of the Diagnostic Interview Schedule (DIS) to distinguish between somatic symptoms of depression that result from physical causes and those that result from psychiatric causes.DESIGN:The data for this study come from the ECA Wave Two surveys sponsored by the National Institute of Mental Health. In this analysis, symptoms of depression the respondent always attributed to physical causes are recoded to be equivalent to those having a psychiatric cause. The third edition of the Diagnostic and Statistical Manual (DSM‐III) scoring algorithm for depression (i.e., dysphoria plus four of the eight Criterion B symptoms) was then applied to the recoded data set.RESULTS:Recoding somatic symptoms of depression, originally attributed to physical or medical explanations, to psychiatric symptoms does not result in a disproportionate rise in diagnosable depression in the older age groups.CONCLUSIONS:This analysis provides additional support for the ECA survey's prevalence estimate of late‐life depression. This research failed to find evidence that the highly structured nature of the DIS makes it unsuitable for ascertaining symptoms of depression in th