The use of a library's collection should be monitored and the results should be taken into consideration in order to establish an effective and efficient collection development and collection management policy. This paper analyzes a total of 884,593 document delivery requests submitted to the National Academic Network and Information Center (ULAKBIM) to identify a core journals collection, and to find out if there is a relationship between the frequency of use of a journal and its impact factor and half-life (obsolescence). A total of 127 core journals satisfied one third of all document delivery requests (both remote and in-house requests). The number of core journals satisfying one third of all requests submitted to the Document Delivery Unit (DDU) and Reference Services Reading Room (RDRR) desk (in-library use) was 184 and 97, respectively, and 87 of which were common. The mean half-life of all journals used in DDU was 10 years while it was 10.3 years for core journals only. The corresponding mean half-life for the same journals calculated on the basis of the number of citations taken from the Journal Citation Reports (2004) was 7.8 years. The mean half-life for all journals used in RDRR was 10 years while it was 7.6 years for core journals only. The corresponding citation-based mean half-life was also 7.6 years. No statistically meaningful relationship was observed between the frequencies of use of core journals and their impact factors. Yet, a low but statistically signifi cant relationship was observed between the use frequencies and total citation counts of RSRR core journals (Pearson's r =.280, p <.01). Findings indicate that the frequencies of the local use of journals are more important than their impact factors and total citation counts when developing and managing a journals collection.