Referential success depends on choice of referring expression. The choice of referring expression will depend on various factors, including speaker and addressee knowledge. A shared-learning paradigm was used in which partners learned names of objects together and separately before a referential task. Items differed on commonality, which were independently rated as common or rare. Speakers were less likely to use names versus other forms when items were rare than common (p<0.001) and less likely to use names when items were new than learned together (p<0.001). Speakers were also more likely to use a name when the addressee was deemed more knowledgeable (p<0.01). We take this to show speakers choose to use a name based on the likelihood that their interlocutor will know the name. Factors affecting the likelihood include prior knowledge of what a typical addressee will know and shared experience, which includes dynamically updated inferences about the interlocutor’s expertise.