Although health is a key element in one's experience of middle adulthood as a time of productivity and personal fulfillment, research on psychosocial factors predictive of mid-life health is sparse, especially for women. Psychosocial variables are not only highly salient to health, but also are potentially modifiable by women themselves. This study employed a multivariate, multitheoretical approach to the study of health, examining a variety of psychosocial predictors (locus of control/mastery, psychological well-being, role quality, social network ties, optimism, and demographic variables) in a secondary analysis of data collected by Baruch and Barnett on 238 women. Subjects were divided into four groups: never married (N=50), married without children (N=54), married with children (N=88), and divorced with children (N=46) and were interviewed in their homes (Brookline, Massachusetts), 1978-79. It was found that 27% of the variance in self-reported health of mothers (whether married or divorced) was accounted for by stress, optimism, employment outside the home, occupational prestige, and quality of experience in the maternal role. Twenty-two percent of the variance in health of married women was explained by stress, quality of experience in the wife role, employment, and occupational prestige. A comparison of the healthiest and the least healthy women revealed that women in better health in middle adulthood had fewer concerns regarding their work, marital roles, and child-rearing roles as compared to their less healthy counterparts. Contains approximately 120 references. (NB)