The cognitive factors of spatial ability, human-computer interaction, problem solving ability, and geographic attitude have been recognized as relevant to teaching and learning GIS. The goal of this research was to examine these cognitive abilities in university students taking an introductory course in GIS; examine any changes in these abilities after completing the class; and examine the relationship between those abilities and the students’ grades in the class. It was hypothesized that students with higher cognitive ability scores would have higher grades than students with lower cognitive ability scores. Nine different self-report surveys were used to assess the students’ spatial, computer, problem solving, and geographic cognitive abilities. The surveys were administered at the beginning and end of the two academic semesters. Analysis of the students’ scores revealed a significant improvement on four of the nine cognitive ability surveys; one that measured computer experience and three that measured spatial ability. Bivariate correlations and multiple regression analyses were used to measure the relationship between the students’ scores on the cognitive ability surveys and the students’ grades. Students received grades on lecture exams, lab exercises, individual projects, and an overall grade. Only two of the bivariate correlations were statistically significant: the factors of geography attitude and learning style were significantly correlated with the students’ project grade. Multiple regression analysis also revealed a very weak relationship, explaining less than 20 percent of the variance between the scores on the cognitive ability surveys and the students’ lecture grade, lab grade, and overall grade. However, a much stronger relationship, explaining more than 45% of the variance, existed between the cognitive ability surveys and the students’ project grade. These findings suggest that cognitive processes utilized for traditional classroom learning to pass lecture exams are different than those utilized to learn the software skills necessary to complete a GIS project. Therefore, it was concluded that the cognitive ability scores are poor predictors of grades related to traditional classroom learning such as lecture exams; however, these scores are more useful as predictors of the grades on a GIS project.