Post-course instructor ratings are a common practice at universities in Europe, Australia, and North America. Rather than solely describing teaching practice, however, such ratings may be associated with a range of non-pedagogical factors. We explored engineering students' instructor ratings at a large United States institution, investigating relationships between overall instructor rating and some of those non-pedagogical factors, including subject area, class size, and students' course grades. Instructor ratings were more favourable in humanities courses than in science or math courses, and students gave higher ratings to instructors who taught smaller class sizes. The strongest relationship existed between overall rating and students' course grades: students who received A's rated instructors an average 0.84 points higher on a 6-point scale than students who received F's. Students who withdrew from a course provided the lowest instructor ratings. When instructor ratings are used as metrics of teaching ability in discussions related to promotion, tenure, or salary adjustments, there is an inherent inequity in the system that must be acknowledged beyond already well-documented biases, as non-pedagogical factors often outside of the instructor's control may be significantly associated with those ratings.