This study examined the unique effects of four variable groups on changes in older adults' depressive symptoms for a 2-year period: (1) baseline health and disability status, (2) changes in health and disability since baseline, (3) stability and changes in marital and caregiving status and in work and volunteering, and (4) stability and changes in health-related behaviors. With data from the 1998 and 2000 interview waves of the Health and Retirement Study, the authors used gender-separate multistep (hierarchical) residualized regression analyses in which the Center for Epidemiological Studies Depression scale (CES-D) score at follow-up is modeled as a function of the effect of each group of independent variables. As hypothesized, changes in health, disability, marital, and caregiving status explained a larger amount of variance than the existing and stable conditions, although each group of variables explained a relatively small amount (0.3-3.4%) of variance in the follow-up CES-D score.