This report describes a use study designed to facilitate the selection of books that will be needed and used by identifying heavily used subjects and publishers in an academic library's collection of nursing books. Discussions include the study background and significance, study limitations, and basic assumptions. In addition, the procedures for data collection are described, including an inventory of the entire collection of 522 nursing books to ascertain their availability for use; the identification of classifications with average circulations higher than the average for the entire collection; and the analysis of publishers to determine those whose books circulated heavily. Based on the findings of this examination, it was considered prudent to purchase all the nursing output of those publishers who most frequently publish nursing books and to purchase duplicates in those classification areas which receive heavy circulation. It was also decided to test the significance of the study findings by monitoring the use of the books purchased as a result of the study and to conduct a user survey to determine user perceptions of collection adequacy. In addition, the use of Spearman rank order correlation to compare sample analysis with complete set analysis indicated that sampling would not be an acceptable technique for examining collections of this size. Statistical results are presented in three tables, and eight references are provided. (KM)