Although people with diabetes mellitus have a high risk of depression and depression may increase mortality among people with other conditions, the impact of depression on mortality risk among people with diabetes needs further examination. Using survival analysis, the authors analyzed longitudinal data from the NHANES I Epidemiologic Follow-up Study (1982-1992). The findings showed that the presence of severe depressive symptoms significantly elevated mortality risk among US adults with diabetes; the same pattern was not observed among people without diabetes. After results were controlled for sociodemographic, lifestyle, and health-status variables, diabetic persons with Centers for Epidemiologic Studies Depression (CES-D) Scale scores of 16 or more had 54% greater mortality than those with scores under 16 (p = 0.004). After exclusion of participants who died during the first year of follow-up, mortality remained higher among those with CES-D scores greater than or equal to 22 as compared with those with CES-D scores less than 16, but not among those with CES-D scores between 16 and 21. No significant relation between depression and mortality was found in the nondiabetic population. This analysis indicates that diabetes modifies the effect of depression on mortality. It also demonstrates the importance of observing subgroups, rather than aggregated populations, when examining the effect of depression on mortality.