A test of factual knowledge of about 10 nations was developed and applied to 96 Oxford children aged from 7 to 11, whose preferences for these nations had already been determined. The relationship between preference for and knowledge about other nations was found to be curvilinear in form and did not seem to be accountable for in terms of a similar relationship in environmentally available cues. Working-class children displayed considerably less knowledge than middle-class children. The causal implications of the findings are discussed.