This study examined whether a set of self-regulated learning variables previously associated with academic achievement could be used to predict first year college student retention. The Motivated Strategies for Learning Questionnaire (MSLQ) was administered to 1,062 first year students with completed questionnaires returned by 677 respondents. After removing outliers, the final sample consisted of 673 cases, with 70% of the cases in the retained group and 30% in the not retained group. The dependent variable was second year retention, and the independent variables were the 15 subscales from the MSLQ. Descriptive statistics and discriminant analysis were used to answer two research questions: (1) Can self-reported self-regulated learning behaviors of first year students discriminate between those students who do and do not persist into their second year of college? and (2) If self-regulated learning behaviors can predict persistence, which factor or combination of factors is best able to discriminate between first year students who do and do not persist into their second year of college? Discriminant analysis indicated that self-regulated learning could discriminate between students who were retained and those who were not. The full model produced a statistically significant discriminant function (F (15, 657) = 1.96, p = .02) that was able to correctly classify 65% of the cases using a linear rule and 62% using a quadratic rule, both rates higher than might be achieved through chance alone (56%). A stepwise discriminant analysis identified a four-factor model that resulted in a statistically significant discriminant function (F(4, 668) = 6.13, p = .00). Using both linear and quadratic rules, the four-factor model was able to improve on the correct classification of retained students compared to the results produced by the full model, correctly assigning 67% of cases. The four-factor model also produced the best combination of classification rates for the two groups when the quadratic rule was applied, correctly assigning 17.82% of the not retained cases and 92.37% of the retained group.