Celeste Calkins, John "Skip" Crooker, and Qingmin Shi (all Decision Support) won the RMAIR 2020 Best Presentation for their recent presentation, “Instructors’ Grading Practices and Student Evaluations of Teaching”, at the annual virtual conference of the Rocky Mountain Association for Institutional Research (RMAIR). This presentation examines how student evaluations of teaching (SET) are related to instructors’ grading practices (grading leniency and grading reliability), instructor characteristics (e.g., gender, ethnicity, rank), student variables (e.g., GPA, gender, academic level), and students’ perceptions about course grade and course challenge. The results reveal that SETs are positively related to the grade a student received and their perception of course challenge and grade. Grading leniency, cumulative semester GPA, faculty gender, ethnicity, and rank also were significant contributors to the model. The preliminary analysis explains more than 64 percent of the variation in individual SET based on the chosen control variables. Controlling for the other factors, students with high cumulative GPAs tend to rate instructors teaching more poorly than their peers with lower cumulative GPAs. Faculty who have relatively high grading leniency tend to be rated lower by students, which goes against much of the current literature. Assigning consistently higher grades does not necessarily imply better evaluation.