Early research on life-stress grappled with the question of whether significant life-events bring about changes in health status. The emphasis has now shifted to the identification of factors that explain why some people seem to be so severely affected by life's adversities and others are not. From a class of what might be called 'vulnerability variables' (Kessler, 1979), support from one's social network has emerged as a significant factor that can account for at least some of the vulnerability differences between groups of stressed individuals. Since Cassel's (1974) review of the evidence linking social upheavals to adverse health consequences for both humans and animals, hundreds of empirical studies have been completed that assess the direct and indirect effects of social support on mental and physical health. This literature is so voluminous as to require several books devoted to reviews of various aspects of it (e.g. Cohen & Syme, 1985; Gottlieb, 1981; and Gottlieb, 1983). In this paper we will distil these as well as highlight some of the recent empirical developments, particularly in those areas that have received less attention in prior reviews. Social support has been defined as the presence of others, or the resources provided by them, prior to, during, and following a stressful event. While there is no general agreement on a single definition, the variety has spawned a number of typologies attempting to organize the literature (e.g. Cohen & Syme, 1985; Cohen & Wills, 1985; Gottlieb, 1983; House & Kahn, 1985). Most of these typologies initially distinguish between functional and structural operationalizations of social support.