The wear-and-tear hypothesis of caregiving (the longer care provided, the more psychological strain on caregivers) was examined using panel survey data from 112 adult children providing interhousehold care to an impaired elderly parent. Measures included subjective caregiving stress and perceived caregiving effectiveness ("wear") and depression and affect balance ("tear"). Contrary to the hypothesis, data revealed variability in children's adaptation to caregiving, improvement rather than deterioration being the norm. In a recursive path model, both subjective stress and perceived effectiveness were significant predictors of changes in depression. Depression was not significantly related to duration of caregiving. Discussion suggests that future researchers consider other types of caregivers and care settings, factors predicting variability in caregivers' adaptation over time, and changes as well as stability in care arrangements.