Background: The positive relation between course grades and student ratings of course satisfaction is well established but controversy continues concerning the magnitude, interpretation, and implications of this association. Aims: This study examined the within course relations of a set of variables often implicated as potential contributors to the grade-student rating relation. Sample: Two-hundred and twenty students enrolled in an internet-based university course. Method: Measures of course mastery and performance goals, expected grades, course effort, cumulative GPA, and final course grade were prospectively collected. The dependent variable was a global measure of course satisfaction obtained at the end of the semester. Results: A path model fit the data very well and all study variables directly or indirectly predicted course satisfaction. Grades had a moderately strong effect, half of which was independent of other model variables. Adjusting for all model variables, students high on mastery course goals evaluated the course more positively whereas students high on performance goals provided less favourable evaluations. Conclusions: Students can assess the same course and instructor in different ways depending upon such factors as their degree of success, their motivations for taking the course, and the amount of effort invested. Course satisfaction, then, can be substantially influenced by factors loosely or unrelated to course or teacher effectiveness.