Cross-sectional studies demonstrate depression is associated with disability in elders. These studies also report that disability in depressed elders is associated with greater medical illness burden, cognitive impairment, and behavioral changes. Only longitudinal studies, however, can determine the impact of depression and its comorbidities on functional decline. This review summarizes the findings of 20 longitudinal studies examining the relationship between baseline or incident depression and functional decline. However, the mediational effects of potential risk factors identified by cross-sectional studies cannot be derived from the current literature. We propose a mediational effects model for future longitudinal studies, incorporating measures sensitive to both mood symptoms and the medical, cognitive, and behavioral comorbidities of depression to better understand the impact of each on functional decline and to focus future clinical interventions.