This chapter deals with coreference puzzles, long regarded as bedrock data against which the adequacy of any theory of singular reference must ultimately be tested. The chapter argues that two-factor referentialism, introduced in Chapter 2, does the best job of solving them. It considers a number of coreference problems, old ones from Frege; newer ones from Kripke, Saul, Richard, and others; and some entirely new ones. It discusses some relevant syntactic issues and distinguishes intrinsic coreference from coincidental coreference. The view that names are individuated by spelling and pronunciation, such as the theory of “nambiguity” put forward by Perry and Korta, is rejected. It criticizes views that treat names as indexicals, as conflating importantly different instruments of reference. It criticizes the view that names are predicates, as having no reasons in its favor other than a misplaced desire for unity. It then develops the two-factor theory by comparing its treatment of cognitive significance with Frege’s views, and more recent views of Perry, Recanati, Fiengo, and May.