Two studies compared representations of the self and of other people, guided by the hypothesis that self-other differences derive from one's greater familiarity with oneself than with others. For the first study, participants wrote open-ended descriptions of themselves, a familiar person, and an unfamiliar person, which were analyzed for the amount and types of information they contained and for consistency in specific content across stimulus people and situations. Participants returned for a second study 1 week later and made timed judgments of information taken from their written protocols. The response latencies for these judgments were used to infer how information is organized in self- and other-concepts. The results supported most of the predicted self-other differences, but almost all were matched by differences between familiar and unfamiliar others. Familiarity does provide a parsimonious explanation for many self-other differences.