In 1990, a study was conducted at Columbia College (CC) to determine whether a predictor variable could be constructed for student nonsuccess in beginning mathematics courses. For the purposes of the study, nonsuccess was defined as earning a grade of D, F, or No Credit either at the time of withdrawal from the course or as a final grade. The ASSET mathematics skills assessment test and a Piaget-based pencil-and-paper test of logical reasoning were administered to 150 students newly enrolled in their first CC mathematics course. Final course grades were collected for 134 of these students. An analysis of the data revealed that a student who obtained "critical scores" on both tests had about a 67% chance of failing the first mathematics course attempted. If only the results of the ASSET test were used, a student who tested "critical" had a 53% chance of failing the first mathematics course. Using both tests offered a better chance of identifying at-risk students. Further statistical analysis suggested that 8% of the variance in grades was accounted for by the ASSET scores alone and that 3% of the variance was accounted for by the Piaget scores alone. Possible explanations for the lack of grade-predictive power of the tests are offered, along with suggestions for further research. (JMC)