The general problem of differential vulnerability of groups to apparently similar stress is addressed in a theory of family vulnerability, based on the proposition that group cohesion may derive from eitheror bothpersonal influence (which is necessarily cohesive and implies regenerative capacity but lack of stability) and positional influence (which may be cohesive and/or coercive and implies efficiency and stability, but lack of regenerative capacity). Taking these as separate variables, ideal types of families and hypotheses of their stress vulnerability are developed, and briefly related to types of stress. Though discussion focuses specifically on family vulnerability, it appears relevant to diverse areas of social and psychological research and theory.