This study examined whether the use of trend analysis combined with analysis of persistence variables can be used to establish a model to forecast the first-year persistence of college freshmen. A linear model was created using data on 2,603 first-time freshmen at a moderate-sized comprehensive university from fall 1989 through fall 1993. The forecast equation was tested on the first-year persistence rate for freshmen newly enrolled in fall 1994. The resulting discrepancy was between the expected and actual persistence rate of 0.56%. The study identified the following variables as having statistically significant correlations with freshman first-year persistence: high school grade point average, high school rank, university grade point average, Scholastic Assessment Test (Math) score, loans, institutional scholarships, outside scholarships, and work-study. Five tables and a figure detail the study's findings. (Contains 12 references.) (Author/LEE)