This study uses a new source of linked census data (N = 6,734) to test theories proposed to explain the high intergenerational coresidence in 19th-century America. Was it a system of support for dependent elderly, or did it reflect intergenerational interdependence? I focus on transitions from middle age to old age, and I assess key predictors of family transitions, including widowhood, retirement, disability, migration, and wealth. The results show that adverse events precipitated changes in the headship of intergenerational families but did not increase the likelihood of residing in an intergenerational family. The findings suggest that 19th-century intergenerational coresidence was not principally a means of old-age support; more often, probably, there was a reciprocal relationship between generations.