We hypothesize that divorce immediately increases psychological distress and has long-term negative consequences for the physical health of divorced people. In addition, we hypothesize that divorce indirectly causes long-term increases in distress through stressful midlife events. The hypotheses are tested using data from 416 rural Iowa women who were interviewed repeatedly in the early 1990s when they were mothers of adolescent children; the women were interviewed again in 2001. The data support the hypotheses. In the years immediately after their divorce (1991-1994), divorced women reported significantly higher levels ofpsychological distress than married women but no differences in physical illness. A decade later (in 2001), the divorced women reported significantly higher levels of illness, even after controlling for age, remarriage, education, income, and prior health. Compared to their married counterparts, divorced women reported higher levels of stressful life events between 1994 and 2000, which led to higher levels of depressive symptoms in 2001.