This paper discusses a multi-dimensional model of ethnocentrism and describes a study that tests the prediction that this multi-dimensional model is a better predictor of communicative behaviors than a single dimensional model of ethnocentrism. First, four theories of ethnocentrism are reviewed in an attempt to find common links among approaches. The first theory discussed is belief congruence--perceptions of similar ingroup-outgroup beliefs, values, and culture. The second theory examined is racial congruence. Some researchers have found that race accounted for more variance in social distance than did belief similarity. Some experimental research on ethnocentrism has been explicitly grounded in the third theory discussed, balance theory, which relies on perceptions of differences to predict ethnocentric behavior. The fourth theory reviewed is the stereotyping/trait-attribution theory. Stereotyping of traits and attributes has been traditionally recognized as central to ethnocentric behavior. The four reviewed theories were found to have two common elements: (1) homophily-heterophily, defined as the degree to which pairs of individuals who interact are similar in certain attributes such as beliefs, values, and education, and social status; and (2) prejudice, an evaluative predisposition to respond to homophily judgments. The argument is thus made that ethnocentrism is a multi-dimensional construct. To test whether this multi-dimensional model of ethnocentrism is a better predictor of communicative behavior than a single dimensional model, 268 students from an introductory class in speech communication from a large university completed various tests. The results of the tests are reported and discussed in the paper. The model received considerable support. (RM)