Australian universities have a long history of use of student satisfaction surveys. Their use has expanded and purpose changed over time. The surveys are often viewed as distorted by external influences such as discipline context, class size and year level of participants. This paper reports on the results of a large-scale investigation interrogating the influence of class size on student satisfaction ratings. The investigation was conducted at a large, comprehensive, research-intensive Australian university. It drew on the data from a survey administered to all students in all units of undergraduate and postgraduate study conducted across the university over four semesters. Data were collated into four class size categories. This categorised data were subjected to statistical analyses. This paper discusses the approach taken, the results of the analysis and their implications for student evaluation of teaching and learning.