Using a ‘near minimal ’ group paradigm (see Taifel et al., 1971), this research tested whether the reasons people think about following group categorization can account for the magnitude of ingroup bias. College students were randomly assigned to four conditions. These conditions instructed subjects either to think about reasons for ingroup choice (‘ingroup’ condition), outgroup choice (‘outgroup’), to think about anything they wanted (‘basic’), or to think about distracting activities (‘distraction’). The hypothesized ordering of ingroup bias and polarized attitudes was: ingroup > basic > distraction > outgroup. The results support both hypotheses. The meaning of these results are discussed in relation to social identity theory and Billig's (1985) rhetorical approach to prejudice.