Mortality-related processes are known to modulate late-life change in cognitive abilities, but it is an open question whether and how precipitous declines with impending death generalize to other domains of functioning. We investigated this notion by using 13-year longitudinal data from now-deceased participants in the Berlin Aging Study (N ϭ 439; 70 -103 years at first occasion; M ϭ 87 years). Using time metrics of chronological age and time-to-death, we compared changes in key indicators of cognitive, sensory, physical, health, social, and self-related domains. Across variables and domains, mortality models revealed steeper average rates of change than age models. However, some domain indicators were more prone to mortality-related change than others. Examining between-person differences, we found that sociodemographic characteristics (surviving to an older age at death, being a woman, lower socioeconomic status) and proxies of pathologies (comorbidities, disability, and suspected dementia) related to lower levels of late-life functioning. In contrast, little evidence was found for correlates of differential change. Our results illustrate both the pervasive nature of progressive processes leading toward death and their domain specificity. Inquiries with more closely spaced multidomain measurements are needed to identify invariant and variable aspects of the end-of-life "cascade."