To learn the meanings of words, children must connect referents in the world around them with the sounds they hear. One proposed mechanism for this process is cross-situational word learning: tracking associations between words and objects across time. We consider the problem of anaphora for a crosssituational word learner: after an object has been introduced it is unlikely to be named in every succeeding reference to it. This problem is particularly pronounced in Japanese, which uses “zero anaphora,” where pronouns can be omitted from utterances. We analyze a corpus of Japanese mothers talking to children about sets of objects, originally recorded by Fernald and Morikawa (1993). Overall rates of anaphora were much higher for Japanese mothers compared with English mothers. Zero anaphora was primarily used when the discourse topic was already established, suggesting that a discourse-finding strategy may be important for word learning in Japanese. In addition, unexpectedly, due to the existence of zero anaphora as a common referential strategy, pronouns were more likely to be used when the topic was new than when it was given (reversing common results for English). Keywords: Child-directed speech; Japanese; zero anaphora; discourse analysis; language acquisition.