The first purpose of this study was to analyze the results of the confirmatory factor analyses, via EQS, with regard to the latent structures of the Learning and Study Strategies Inventory (LASSI) (C. Weinstein, D. Palmer, and A. Schulte, 1987) as proposed by S. Olejnik and S. Nist (1992), A. Olivarez and M. Tallent-Runnels (1994), B. Olaussen and I. Braten (1998), and the model that evolved from the exploratory factor analysis of the current data set. Data were from two samples, one of 319 college freshmen and the other of 2,535 college freshmen. The aim was not only to determine the model that best fits the data set, but also to investigate how replicable these models are with a large sample size. The model proposed by Olaussen and Braten appears to have the best fit to the data of this study. The two-factor model that evolved from the exploratory factor analysis of the data did not adequately explain the data under the scrutiny of confirmatory factor analysis. This study supports the three latent structure model of the LASSI instrument. The second purpose of the study was to examine the reliability (internal consistency) of the 10 LASSI subscales. The results show that, in some subscales, the Cronbach alpha obtained from this study appear to be slightly better than the indices reported in the manual. Overall, the alpha coefficients obtained in this study reasonably matched those reported in the LASSI manual. (Contains 2 tables, 4 figures, and 16 references.) (Author/SLD)