Student evaluations of teaching (SETs) are widely used to assess instructors and courses in higher education. Previous research has identified patterns of gender bias in SETs, yet no such bias was found in previous internal reports at the University of Amsterdam (UvA). The present study replicated UvA’s previous investigations using current SET data and extended them through the use of Bayesian multilevel modeling. Several models were estimated to assess whether instructor gender predicted SET ratings and whether specific evaluation items were more prone to gender bias. The models tested gender differences in composite SET scores, variation across faculties, and the role of course-level covariates. Overall, the findings provided evidence for the absence of meaningful gender bias in UvA’s SETs.