This study tested the hypotheses that the perceived magnitude and diversity of an opinion group are enhanced by membership of that group, and that perceptions of the magnitude and diversity of opinion groups are not confounded with each other. Subjects were asked (i) to indicate their own opinions on each of four issues, and (ii) to estimate the proportion of their peers who would fall into each of three specified opinion groups on each issue, and rate the diversity of membership of each opinion group. The order in which these questions were posed was reversed for half the subjects. Results were consistent with the magnitude hypothesis, but there was little support for the diversity hypothesis. Correlations between magnitude and diversity ratings tended to be positive and reliable. Question sequence exerted no consistently reliable effects. Discrepancies between present and previous findings are discussed.