Social values refer to stable preferences for certain patterns of outcomes to oneself and others (Mc-Clintock, 1972, 1978. Recent research indicates that people with different social values perceive the cooperative-competitive distinction differently. Cooperators see it as an evaluative, good-bad dimension, whereas noncooperators see it as a dynamic, strong-weak dimension (Liebrand, Jansen, Ryken, & Suhre, 1986). We found behavioral support for this principle in that cooperators and competitors displayed different egocentric biases. In Study 1 we expected cooperators to display an egocentric bias by associating themselves with good (cooperative) behaviors and by associating others with bad (competitive) behaviors. Noncooperators, not perceiving cooperation-competition as evaluative, should not display this association as strongly. We had subjects write lists of cooperative and competitive behaviors that either they or others did. Cooperators showed an egocentric bias; competitors did not. In Study 2 a different group of subjects rated a subset of behaviors produced in Study 1. When using an evaluative (good-bad) scale, the behaviors written by cooperators were rated more extremely than the behaviors written by competitors. Contrary to expectations, when using a potency (strong-weak) dimension, behaviors written by competitors were not rated more extremely than the behaviors written by cooperators.Social values refer to stable preferences for certain patterns of outcomes to oneself and others (McClintock, 1972(McClintock, , 1978.Through the use of a measurement procedure involving decomposed games (e.g., Messick & McClintock, 1968), individuals can be classified into distinct groups on the basis of those preferences. The three values that have received the most empirical and theoretical attention are (a) cooperation, characterized by a preference to maximize the joint outcome between two people; (b) competition, maximizing one's relative advantage against another person; and (c) individualism, the tendency to maximize one's own outcome without regard to the outcomes of others (e.g., Bern & Lord, 1979).Previous research has shown that actors' social values can influence behavior in two-person mixed-motive game settings