Attraction and hostility find expression in almost every variety of human relationship, and have consequently provided a central theme for social psychology since its beginnings. Yet attempts to conceptualize the diverse phenomena embraced in these terms have produced theories of such wide generality that they have little explanatory or predictive force. The object of the present study is to bring precision to a vast and sprawling area by setting limits and dimensions to the phenomena and investigating them experimentally on the basis of a series of hypotheses derived from a critical analysis of current conceptual approaches, including frustration, need-satisfaction, and dissonance models.The programme of experimental studies focuses on cognitive validation-a motivation to form and maintain subjectively valid evaluations of the self and the social environment-which is shown to be a common denominator of a number of attraction and hostility measures. The results throw light on reactions to boastfulness and to self-debasement; impressions of persons who are described by biased informants; effects of self-evaluation on competitiveness, and the projection of unfavourable characteristics.The interest of the study for social psychologists derives both from its theoretical integration of a wide range of behaviour and from its contribution to experimental design.