MCA has aspired to underscore the significant influence categorization can have on the way members of a culture experience their social reality and assume their roles in it. By following such aspirations, analysts inevitably run the risk of projecting "common-sense" assumptions onto their data and having their work characterized as "wild and promiscuous" (Stoke, 2012). So as to avoid these pitfalls, analysts have searched for ways of making assertions based on the data alone, by treating membership categorization as analytically pertinent only when it is demonstrably relevant and procedurally consequential to the interaction. The trouble with this approach, however, is that the relevance of a certain category might not always be obvious in a single turn or even a single segment of talk. Categories and category-bound predicates may be unpacked throughout an entire interaction and explicitly named only in passing, if at all. This paper aims to show how one category--that of 'girls/women'--could be demonstrated as relevant via the participants' invocation of and orientation to other, connected categories and category-bound activities. The analysis shows that participants build a shared understanding of categories throughout their conversation, and that the demonstrable relevance and procedural consequentiality of these categories becomes evident only gradually as the interaction progresses. Almost an entire hour before they explicitly mention it, the participants seem to orient to the 'girls/women' category by invoking the related category of 'men/guys/boys' and its category-bound predicates. They seem to treat the 'men' and 'women' categories as a standardized relational pair (as having duties and moral obligations to one another), and, by invoking the 'men' category, the speakers appear to make potentially relevant their (unstated) orientation to and membership in the 'women' category.