The goal of this project is to investigate health and mortality differences by marital status among the elderly U.S. population. Numerous studies have established that married persons enjoy better health and lower mortality risks than do unmarried persons of equal age. Much less is known about the sources of these differentials. The specific aims of the proposed analysis are to use the 1984 to 1990 Longitudinal Study of Aging (LSOA) to determine whether these marital status effects result largely from processes which operate at younger ages or whether these processes also operate in later life; to investigate the extent to which the health and mortality differences at the older ages can be explained by differences in the social and economic environment between married and unmarried persons; and to explore gender variation in marital status effects. Data contained in the LSOA are well-suited to the study of marital status differentials in health and mortality among the elderly. A particular advantage of this survey is the collection of repeated measures of marital status, economic status, social support, and physical functioning that will provide more complete and accurate knowledge about the relationships between social and economic factors and health and mortality among the elderly. The LSOA, through its link to the National Health Interview Survey Supplement on Aging, contains detailed information about baseline health status that is unavailable from the data sources used in previous studies. Controls for baseline health status are essential for identifying attributes that continue to influence health at older ages. The results of the proposed study will be important both to researchers in the social sciences and to the policy community. The proportion of elderly who live alone has been steadily increasing and the proportion of older persons who are unmarried is projected to increase substantially over the next 50 years. Of primary concern are implications for the provision of care, especially long-term care for the elderly. Interventions to provide or encourage support for elderly persons living in the community as an alternative to institutionalization may be promoted for financial or humanitarian reasons. Research indicating that social support can act to increase longevity and health among the elderly provides additional compelling rationales for these interventions. Formulation of appropriate policies requires identification of the specific aspects of the social and economic environment that affect the health and longevity of older persons.