In the summer of 1984, seven Library of Congress subclasses in the Columbia-Greene Community College Library were surveyed to determine if past circulation records could reliably predict future circulation. In the summer of 1985, a follow-up study surveyed 10 Library of Congress subclasses to verify the results of the first study. The second study, described in this report, included 3,391 "old" books (acquired in 1976 or earlier) and 1,009 "new" books (acquired in 1977 or 1978). The sample was approximately 10% of the collection. The study confirmed the findings of the first study: the likelihood of books being used does not progressively decline with additional years of non-use. Also reaffirmed is the finding that, while first use could not be reliably predicted, prior non-use can be useful as a predictor for future use, and therefore, as a criterion for weeding. Thirty-two charts accompany this report. (THC)