This article adapts the assumptions of the continuum model of impression formation to address the development of shared impressions in dyads. Two experiments tested the additional assumption that the development of shared impressions depends on consensus motivation. Consensus motivation should focus conversation around stereotype-consistent information and encourage shared stereotypic impressions. However, when dyads are motivated by both consensus and accuracy, they should increase focus on stereotype-inconsistent information, then form shared individuated impressions. In both experiments, hypotheses generally were confirmed regarding consensus motivation. However, hypotheses regarding accuracy motivation met with mixed success; individuating processes were detected but failed to encourage individuated impressions. The discussion explores reasons why these experiments provide only partial support for the extended continuum model.