Objective-Determine the type of online chat questions to help inform staffing decisions for chat reference service considering their library's service mandate. Design-Content analysis of consortial online chat questions. Setting-Large academic library in Canada. Subjects-Analysis included 2,734 chat question transcripts. Methods-The authors analyzed chat question transcripts from patrons at the institution for the period of time from September 2013 to August 2014. The authors coded transcripts by question type using a coding tool created by the authors. For transcripts that fit more than one question type, the authors chose the most prominent type. Main Results-The authors coded the chat questions as follows: service (51%), reference (25%), citation (9%), technology (7%), and miscellaneous (8%). The majority of service questions were informational, followed by account related questions. Most of the reference chat questions were ready reference with only 16% (4% of the total number of chat questions) being in-depth. After removing miscellaneous questions, those that required a high level of expertise (in-depth reference, instructional, copyright, or citation) equaled 19%. Conclusion-At this institution, one in five chat questions needed a high level of expertise. Library assistants with sufficient expertise could effectively answer circulation and general reference questions. With training they could triage complex questions.