In the past decade, a body of literature has evolved around the concept caregiving the (Brody, 1981, 1990). Women midlife who are primary caregivers for older relatives are defined as in the of stressful lifestyles that may include employment, housework, and care of dependent children (Brody, 1981, 1990; Horowitz, 1985; Noelker & Wallace, 1985; Soldo & Myllyluoma, 1983; Stueve & O'Donnell, 1989). Researchers, practitioners, and lay people acknowledge that caregiving is stressful (George, 1990; Zarit, 1989). Further, research continually shows that most primary caregivers to older people are women (e.g., Coward & Dwyer, 1990; Stone, Cafferata, & Sangl, 1987). The degree to which this situation is problematic for caregiving women may not be as grave as some have argued (Spitze & Logan, 1990). There are several reasons--demographic and methodological--for questioning the extent of the problem of women the middle. First, some research has concentrated on whether the situation of women the middle is normative the general population (Spitze & Logan, 1990). These authors argue that the situation defined as women (or men) the middle is not typical of middle-aged persons the United States. That is, the situation of women being the middle is not normative. Similar findings have been reported for middle-aged women Canada (Rosenthal, Matthews, & Marshall, 1989). These authors find, for example, that when women are caring for frail parents, they and their parents tend to be older. Therefore, only a small proportion of women who are providing care to older relatives also have dependent children. Second, several aspects of caregiving women's lifestyles have not been directly examined. For example, there is an assumption that caregiving for dependent parents or children is always stressful, even though comparisons of caregiving and noncaregiving populations are generally not made (George & Gwyther, 1986). Research shows that more caregiving demands (i.e., greater elder impairment, more caregiving tasks performed) are associated with greater caregiver burden or strain (see George, 1990). But how strained is strained? That is, demonstrating a positive, linear relationship between, say, elder impairment and caregiver strain does not indicate the degree to which people are strained, only that as elder impairment increases, caregiver strain also increases. Some researchers suggest that the competing demands of parent care, child care, and employment lead to greater levels of caregiver stress (Older Women's League, 1989; Stone, Cafferata, & Sangl, 1987). The point usually made is that a higher number of demanding situations is more stressful. Because certain proportion of women are caring for older relatives and have dependent children and/or are employed, it is thought that they must be experiencing greater stress because of these multiple or competing demands. The link between these multiple roles and stress is assumed, not demonstrated. Third, some researchers find that care of children exacerbates the stress for women the middle (Rosenthal, Matthews, & Marshall, 1989). Others show that caregivers whose household size is largest describe a pattern of increasing support over time (Clipp & George, 199O, p. S-106). Yet, the relationship between the presence or number of children within a household and caregiver stress or well-being has not been analyzed. The type of living arrangement (shared versus separate), however, has been found to be a important factor affecting caregiver strain (Soldo & Myllyluoma, 1983; Tennstedt & McKinlay, 1989). Adult children caregivers sharing the same residence with care recipients report greater strain or depression than those who live apart from the recipient. Often, the elders' greater impairment prompts the shared living arrangement. In other cases, the increased contact shared living arrangements may lead to increased strain Noelker & Wallace, 1985). …