A large data set of student ratings of college teaching was studied to explore previous indications that student ratings are affected by the subject area of instruction. Previous research by Marsh (1984), Feldman (1978), and Cranton and Smith (1986) reported the influence of factors other than instruction. To examine the impact of subject area on ratings, data were analyzed from the IDEA (Instructional Development and Effectiveness Assessment) System Survey Form. Ratings were examined in 44 subject areas from 87,843 college classes. Phase I focused on three summary outcome items: (1) overall instruction; (2) desire to take another course from this instructor; and (3) improved attitude toward course of study. Results were generally similar to Feldman's; economics received the lowest ratings, followed by mathematics and statistics, physics, computer and information studies, and engineering. Higher ratings were reported for fine and applied arts, and nursing. Phase II manipulated the sample to control for sources of contamination; results supported Feldman's patterns. Humanities and arts received higher ratings than social sciences, which received higher ratings than mathematics and science. (Explanations are suggested and a copy of the IDEA Survey Form is appended.) (MAC)