Abstract Objective: The main aim of this study is to add new evidence on the descriptive epidemiology of depressed mood, and to investigate suspected determinants for depressed mood in adulthood. Methods: The data are from a continuing survey of a nationally representative sample of adult household residents in the United States, conducted in 1995 and 1996, totaling 26,883 respondents. Multiple logistic regression procedures yielded estimated associations. Results: We found that an estimated 1.9% of adult females and 1.0% of adult males experience a spell of sustained depressed mood during a span of ∼2 weeks duration (i.e. point prevalence). For most of these cases, this is not the first spell. Among women, the smoothed curve for the prevalence estimates shows a peak in the youngest age stratum and decreases across age strata before 60 years, and has a slight secondary peak thereafter. In contrast, for males, the prevalence estimates of depression show no peak in the older age strata. Evidence from logistic regression analyses supports the inference of this later life peak in frequency of depressed mood among women. These new findings add to a growing body of epidemiological evidence on age and depression, and provoke new questions about the possibly etiological relationships involving social structural characteristics of local neighborhoods in combination with individual-level risk factors that have received primary attention in recent psychiatric epidemiology. Conclusions: These findings point to the need for further etiological research, including studies of relationships between social structural characteristics of local neighborhoods and the occurrence of spells of depressed mood, as well as clinical implications for depression mood in late life.