Differences in speaking time during a group interaction were hypothesized to reflect differences in individual dominance. In order to test this assumption, a meta-analysis was conducted. Whether speaking time is used to convey dominance to the same extent that it is used in inferring dominance was tested by contrasting studies concerned with dominance expressed in speaking time with studies of inferred dominance based on speaking time. The relationship between dominance and speaking time was significant. The strength of the associations differed due to the influence of moderator variables. Results showed that inferred dominance studies showed stronger associations between speaking time and dominance as opposed to the expressed dominance studies. If dominance was expressed due to dominance-role assignments, the association between speaking time and dominance was stronger than if individuals with different levels of trait dominance interacted. For men, the association between speaking time and dominance was stronger than for women, and same-gender groups showed stronger associations than opposite-gender groups. Also, increasing group size intensified the strength of the association linearly.